• Re: Suspension losses

    From Roger Merriman@21:1/5 to Frank Krygowski on Thu Jan 2 11:45:31 2025
    Frank Krygowski <frkrygow@sbcglobal.net> wrote:
    Good article from Jan Heine on benefits of wider, softer tires for
    absorbing vibration and lessening suspension losses:

    https://www.renehersecycles.com/the-missing-link-suspension-losses/

    At the time the rumble strip test was published, I expressed some
    skepticism because its roughness is fundamentally different than the
    random roughness of either a rough road or a gravel road. In particular,
    the rumble strip is all "negative," cut into the smooth surface, while
    rough or gravel roads have both "negative" holes plus "positive" patches
    or rocks that protrude above the surface. One practical difference is
    that when dealing with only "negative" roughness, higher speeds reduce losses. The opposite is true with "positive" roughness.

    But I suppose for demonstrating the fundamental effect, the consistency
    of the rumbles is useful. And the measurements seem valid as long as the
    test speed is also consistent.

    Not that convinced to be honest, for a starters folks aren’t going to be riding rumble strips but by mistake!

    And if you’re going to be real world testing, testing on dirt roads with
    all of the inconsistencies that brings is what gravel riders do. With the
    dips as well as the bumps, plus ruts etc.

    Rumble strip testing seems somewhat misleading ie it’s not that controlled nor what riders do.

    As ever claims that they influence pro athletes etc and started the wider
    tire use, IMO it along with disks was adapted by consumers/commuters with
    pro racers lagging behind with adoption and haven’t gone quite as wide, ie stopped at 28 for the Pros where as 30/32 are fairly common among club
    riders.


    BTW, Jobst Brandt is mentioned in the article. I recall that in
    discussing rolling resistance here, he insisted that "rolling
    resistance" should be defined _only_ as the losses generated by tire
    rubber's hysteresis. I disagreed, because that implied that solid rubber tires a la 1880, or near infinite tire pressure, or even metal rims with
    no tire, would be best. Anyone who has ridden an antique solid tire
    "safety" bike knows how slow those tires were.


    As ever is a what you want as well, on the old school road bike, I commute
    on, 28mm felt on the twitchy side 32mm much more planted, the speed
    difference I’m less concerned about, though at that level maybe wider is faster? What is faster would depend on road/bike/rider.

    Roger Merriman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Roger Merriman on Thu Jan 2 07:35:15 2025
    On 1/2/2025 5:45 AM, Roger Merriman wrote:
    Frank Krygowski <frkrygow@sbcglobal.net> wrote:
    Good article from Jan Heine on benefits of wider, softer tires for
    absorbing vibration and lessening suspension losses:

    https://www.renehersecycles.com/the-missing-link-suspension-losses/

    At the time the rumble strip test was published, I expressed some
    skepticism because its roughness is fundamentally different than the
    random roughness of either a rough road or a gravel road. In particular,
    the rumble strip is all "negative," cut into the smooth surface, while
    rough or gravel roads have both "negative" holes plus "positive" patches
    or rocks that protrude above the surface. One practical difference is
    that when dealing with only "negative" roughness, higher speeds reduce
    losses. The opposite is true with "positive" roughness.

    But I suppose for demonstrating the fundamental effect, the consistency
    of the rumbles is useful. And the measurements seem valid as long as the
    test speed is also consistent.

    Not that convinced to be honest, for a starters folks aren’t going to be riding rumble strips but by mistake!

    And if you’re going to be real world testing, testing on dirt roads with all of the inconsistencies that brings is what gravel riders do. With the dips as well as the bumps, plus ruts etc.

    Rumble strip testing seems somewhat misleading ie it’s not that controlled nor what riders do.

    As ever claims that they influence pro athletes etc and started the wider tire use, IMO it along with disks was adapted by consumers/commuters with
    pro racers lagging behind with adoption and haven’t gone quite as wide, ie stopped at 28 for the Pros where as 30/32 are fairly common among club riders.


    BTW, Jobst Brandt is mentioned in the article. I recall that in
    discussing rolling resistance here, he insisted that "rolling
    resistance" should be defined _only_ as the losses generated by tire
    rubber's hysteresis. I disagreed, because that implied that solid rubber
    tires a la 1880, or near infinite tire pressure, or even metal rims with
    no tire, would be best. Anyone who has ridden an antique solid tire
    "safety" bike knows how slow those tires were.


    As ever is a what you want as well, on the old school road bike, I commute on, 28mm felt on the twitchy side 32mm much more planted, the speed difference I’m less concerned about, though at that level maybe wider is faster? What is faster would depend on road/bike/rider.

    Roger Merriman



    I don't have a coherent argument either way but a rumble
    strip test introduces a repeatable experience so that
    various data may be compared. Each rider on a dirt or
    gravel path, and each ride experience by any given rider, is
    an unique set of impedimenta such that data cannot be as
    readily compared.

    In short you make an interesting point but it's not
    measurable for comparative purposes.

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to bp@www.zefox.net on Thu Jan 2 10:06:08 2025
    On 1/2/2025 9:42 AM, bp@www.zefox.net wrote:
    AMuzi <am@yellowjersey.org> wrote:

    I don't have a coherent argument either way but a rumble
    strip test introduces a repeatable experience so that
    various data may be compared. Each rider on a dirt or
    gravel path, and each ride experience by any given rider, is
    an unique set of impedimenta such that data cannot be as
    readily compared.

    AMuzi <am@yellowjersey.org> wrote:

    I don't have a coherent argument either way but a rumble
    strip test introduces a repeatable experience so that
    various data may be compared. Each rider on a dirt or
    gravel path, and each ride experience by any given rider, is
    an unique set of impedimenta such that data cannot be as
    readily compared.

    But one can observe that in the case of smooth pavement,
    suspension losses vanish, while hysteresis losses persist.

    In the end a bike is an overdamped resonator excited by the
    pavement and damped by hysteresis, separately in the tire and
    suspenesion. In that limit, suspension would be faster if used
    with very hard tires on very smooth surfaces. In the limit of
    hard tires and no suspension, the dissipative element becomes
    the rider whose elastic properties are apt to be poor, perhaps
    accounting for the apparent slowness of solid tires.

    Use of a rumble strip for testing is equivalent to selecting
    a particular excitation spectrum. Choice of spectrum will affect
    dissipation depending on internal resonances of the bike/rider
    system. A real road likely corresponds to a 1/f spectrum, but
    a rumble strip will likely be something else. How much difference
    that makes isn't clear but it could be estimated using a mechanical
    analogy equivalent circuit of the kind used to model loudspeakers.

    A pair of series RLC circuits (one for the road-tire interface
    and a second for the suspension-rider interface) would be a good
    start. I'm not skilled enough to do the calculations, but others
    on this group likely are. The hardest part is apt to be finding
    an equivalent circuit for the rider, who isn't a rigid mass but
    rather a dissipative blob....8-)

    Thanks for reading,

    bob prohaska


    Clever.

    I take from that, you think the actual impact/height
    change/velocity change etc from various irregular surfaces
    can be quantified for any given random gravel (or road)
    experience and used to compare efficiency for other iterations.

    I hadn't thought of that, but if that's true then the rumble
    strip test isn't necessary for comparison. Which assumes
    sensors have adequate sensitivity across whatever range and
    that software for that data truly derives actual impedimenta
    values.

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From bp@www.zefox.net@21:1/5 to AMuzi on Thu Jan 2 15:42:31 2025
    AMuzi <am@yellowjersey.org> wrote:

    I don't have a coherent argument either way but a rumble
    strip test introduces a repeatable experience so that
    various data may be compared. Each rider on a dirt or
    gravel path, and each ride experience by any given rider, is
    an unique set of impedimenta such that data cannot be as
    readily compared.

    AMuzi <am@yellowjersey.org> wrote:

    I don't have a coherent argument either way but a rumble
    strip test introduces a repeatable experience so that
    various data may be compared. Each rider on a dirt or
    gravel path, and each ride experience by any given rider, is
    an unique set of impedimenta such that data cannot be as
    readily compared.

    But one can observe that in the case of smooth pavement,
    suspension losses vanish, while hysteresis losses persist.

    In the end a bike is an overdamped resonator excited by the
    pavement and damped by hysteresis, separately in the tire and
    suspenesion. In that limit, suspension would be faster if used
    with very hard tires on very smooth surfaces. In the limit of
    hard tires and no suspension, the dissipative element becomes
    the rider whose elastic properties are apt to be poor, perhaps
    accounting for the apparent slowness of solid tires.

    Use of a rumble strip for testing is equivalent to selecting
    a particular excitation spectrum. Choice of spectrum will affect
    dissipation depending on internal resonances of the bike/rider
    system. A real road likely corresponds to a 1/f spectrum, but
    a rumble strip will likely be something else. How much difference
    that makes isn't clear but it could be estimated using a mechanical
    analogy equivalent circuit of the kind used to model loudspeakers.

    A pair of series RLC circuits (one for the road-tire interface
    and a second for the suspension-rider interface) would be a good
    start. I'm not skilled enough to do the calculations, but others
    on this group likely are. The hardest part is apt to be finding
    an equivalent circuit for the rider, who isn't a rigid mass but
    rather a dissipative blob....8-)

    Thanks for reading,

    bob prohaska

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From bp@www.zefox.net@21:1/5 to AMuzi on Thu Jan 2 17:06:09 2025
    AMuzi <am@yellowjersey.org> wrote:

    I take from that, you think the actual impact/height
    change/velocity change etc from various irregular surfaces
    can be quantified for any given random gravel (or road)
    experience and used to compare efficiency for other iterations.

    To a decent approximation, yes. Any surface profile can be represented
    by a spectrum. Music is usually represented by "pink" noise, thermal
    noise is white (uniform) and most real systems have noise proportional
    to 1/f (imagine turn-on transients as singularities). A real road would
    likely be some combination with the radius of the tire serving as a high frequency filter.

    I hadn't thought of that, but if that's true then the rumble
    strip test isn't necessary for comparison. Which assumes
    sensors have adequate sensitivity across whatever range and
    that software for that data truly derives actual impedimenta
    values.

    One would have to measure the force/deflection curves for both tires
    and suspension elements, along with the masses of the sprung and unsprung elements. Since losses are rate dependent, especially for suspensions
    with hydraulic damping, a range of speeds/frequencies would have to be measured. I think an accurate model would get fairly complicated, especially
    if the rider were included. Each compliance (tire, suspension spring, seat spring and rider body part that deflects) would have to be accounted for.

    There are potentially four coupled resonators: Tires, swingarm/forks and finally rider (divided into arms/torso sections probably). Overall, tests
    on a rumble strip or drum with some kind of ergometer might be simpler.

    Very likely the motor racing industry already has software that can do the analysis. Most of the interest in that market is controlling resonances,
    not minimizing losses, but otherwise the problems are very similar.

    Thanks for writing!

    bob prohaska

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Frank Krygowski on Thu Jan 2 11:35:05 2025
    On 1/2/2025 11:17 AM, Frank Krygowski wrote:
    On 1/2/2025 11:06 AM, AMuzi wrote:
    On 1/2/2025 9:42 AM, bp@www.zefox.net wrote:

    But one can observe that in the case of smooth pavement,
    suspension losses vanish, while hysteresis losses persist.

    I think that would be true only if the smooth pavement were
    as smooth as a linoleum floor. Or a wooden track. IIRC, what
    got Jan Heine started on investigations of rolling
    resistance vs. tire width was coast-down tests on a Soapbox
    Derby track. I suspect that was quite smooth. Soapbox cars
    have hard tires and no suspension, AFAIK.

    In the end a bike is an overdamped resonator excited by the
    pavement and damped by hysteresis, separately in the tire
    and
    suspenesion. In that limit, suspension would be faster if
    used
    with very hard tires on very smooth surfaces. In the
    limit of
    hard tires and no suspension, the dissipative element
    becomes
    the rider whose elastic properties are apt to be poor,
    perhaps
    accounting for the apparent slowness of solid tires.

    Use of a rumble strip for testing is equivalent to selecting
    a particular excitation spectrum. Choice of spectrum will
    affect
    dissipation depending on internal resonances of the bike/
    rider
    system. A real road likely corresponds to a 1/f spectrum,
    but
    a rumble strip will likely be something else. How much
    difference
    that makes isn't clear but it could be estimated using a
    mechanical
    analogy equivalent circuit of the kind used to model
    loudspeakers.

    A pair of series RLC circuits (one for the road-tire
    interface
    and a second for the suspension-rider interface) would be
    a good
    start. I'm not skilled enough to do the calculations, but
    others
    on this group likely are.

    I _may_ have been able to do such calculations 50 years ago,
    but I'm not sure. I certainly can't do them now.

    The hardest part is apt to be finding
    an equivalent circuit for the rider, who isn't a rigid
    mass but
    rather a dissipative blob....8-)

    I actually think physically modeling that dissipative blob
    might be valuable for the tire industry. Using such a blob
    to apply weight during a rolling drum test might give better
    information than what those tests give now.


    Clever.

    I take from that, you think the actual impact/height
    change/velocity change etc from various irregular surfaces
    can be quantified for any given random gravel (or road)
    experience and used to compare efficiency for other
    iterations.

    I hadn't thought of that, but if that's true then the
    rumble strip test isn't necessary for comparison. Which
    assumes sensors have adequate sensitivity across whatever
    range and that software for that data truly derives actual
    impedimenta values.

    There are ways of quantifying roughness, with varying
    scales, varying tools. I'm most familiar with roughness
    measurement of machined parts, with tools varying from
    sample cards for "fingernail" test comparisons, to RMS
    readers akin to phonograph needles or laser scattering devices.

    https://en.wikipedia.org/wiki/Surface_roughness

    ISTR reading about systems for evaluating pavement fairly
    crudely, as in whether it should be repaved or not. I don't
    know of a system actually used for measuring pavement
    roughness at a scale affecting bike tire choice.


    Yes, I'm familiar with surface finish (roughness) numbers in
    machining, but an offroad bicycle, for example on a gravel
    path (bianca strada) or babyheads (much of Paris Roubaix)
    would be a series of variable impedimenta in some chaotic
    non-order for height & frequency. The principle is the same
    but the amount of data is staggering.

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Roger Merriman@21:1/5 to Frank Krygowski on Thu Jan 2 19:34:57 2025
    Frank Krygowski <frkrygow@sbcglobal.net> wrote:
    On 1/2/2025 12:06 PM, bp@www.zefox.net wrote:
    AMuzi <am@yellowjersey.org> wrote:

    I take from that, you think the actual impact/height
    change/velocity change etc from various irregular surfaces
    can be quantified for any given random gravel (or road)
    experience and used to compare efficiency for other iterations.

    To a decent approximation, yes. Any surface profile can be represented
    by a spectrum. Music is usually represented by "pink" noise, thermal
    noise is white (uniform) and most real systems have noise proportional
    to 1/f (imagine turn-on transients as singularities). A real road would
    likely be some combination with the radius of the tire serving as a high
    frequency filter.

    I hadn't thought of that, but if that's true then the rumble
    strip test isn't necessary for comparison. Which assumes
    sensors have adequate sensitivity across whatever range and
    that software for that data truly derives actual impedimenta
    values.

    One would have to measure the force/deflection curves for both tires
    and suspension elements, along with the masses of the sprung and unsprung
    elements. Since losses are rate dependent, especially for suspensions
    with hydraulic damping, a range of speeds/frequencies would have to be
    measured. I think an accurate model would get fairly complicated, especially >> if the rider were included. Each compliance (tire, suspension spring, seat >> spring and rider body part that deflects) would have to be accounted for.

    There are potentially four coupled resonators: Tires, swingarm/forks and
    finally rider (divided into arms/torso sections probably). Overall, tests
    on a rumble strip or drum with some kind of ergometer might be simpler.

    Very likely the motor racing industry already has software that can do the >> analysis. Most of the interest in that market is controlling resonances,
    not minimizing losses, but otherwise the problems are very similar.

    One further thought: If we accept (as I do) that jiggling the human
    pedaler does cause loss in energy and speed, why aren't we all using
    saddles with some sort of damped springing?

    I know suspension seatposts exist, but even those are not popular on
    road bikes.

    Is some on gravel bikes and stem’s which tend to be fairly basic and move
    in less than ideal ways, ie the stems tend to pivot which some riders
    really don’t like.

    Also while lighter they are outperformed by suspension ie Gravel suspension forks.

    But probably mostly that roadies and most Gravel riders are, are as roadies
    are ie fairly conservative with technology choices.

    ISTM that if more "suspension" is valuable via wider tires, it might
    also be valuable via sprung saddles, if done right.

    There is a difference between the bike being sprung and the rider,
    something Specialised talked about with the Diverge and specifically the
    one with suspension in the frame as well as the fork.

    Which gives a different feel to for example suspension such as MTB has even reduced down to gravel travel. Plus performance as well clearly.

    My wife used to ride a Brooks B72. Its four curly support wires gives
    just a bit of spring action. It's now on my about-town 3 speed. That
    bike never goes far, but I don't detect any detriments to the slight springiness.

    Its not really a saddle intended for more than that, and for that use-case, probably more that roadie and even MTBers are unlikely to like the feel of
    the saddle moving in relation to the bottom bracket even if mildly.

    Roger Merriman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From bp@www.zefox.net@21:1/5 to Frank Krygowski on Fri Jan 3 01:45:49 2025
    Frank Krygowski <frkrygow@sbcglobal.net> wrote:

    One further thought: If we accept (as I do) that jiggling the human
    pedaler does cause loss in energy and speed, why aren't we all using
    saddles with some sort of damped springing?

    I know suspension seatposts exist, but even those are not popular on
    road bikes.

    I'm using a suspension seatpost now, removed from a town bike. It's
    slightly more comfortable. No idea if it's more efficient. Certainly
    heavier, probably lossy unless I balance pedal effort to keep pressure
    on the saddle constant. That difference is small at most.

    Thanks for writing,

    bob prohaska




    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From bp@www.zefox.net@21:1/5 to Frank Krygowski on Fri Jan 3 02:03:51 2025
    Frank Krygowski <frkrygow@sbcglobal.net> wrote:
    On 1/2/2025 11:06 AM, AMuzi wrote:
    On 1/2/2025 9:42 AM, bp@www.zefox.net wrote:

    But one can observe that in the case of smooth pavement,
    suspension losses vanish, while hysteresis losses persist.

    I think that would be true only if the smooth pavement were as smooth as
    a linoleum floor.

    To first principles, it's the compliance of the tire that keeps alternating forces out of the suspension. If suspension doesn't deflect there can't be
    any losses. Since the tire always deflects it's always lossy. On a very
    hard (10 bar) tire, that might take a linoleum floor. But, it's the tire
    that decides what's "smooth".

    Thanks for writing,

    bob prohaska

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From bp@www.zefox.net@21:1/5 to AMuzi on Fri Jan 3 02:26:32 2025
    AMuzi <am@yellowjersey.org> wrote:

    Yes, I'm familiar with surface finish (roughness) numbers in
    machining, but an offroad bicycle, for example on a gravel
    path (bianca strada) or babyheads (much of Paris Roubaix)
    would be a series of variable impedimenta in some chaotic
    non-order for height & frequency. The principle is the same
    but the amount of data is staggering.


    But, what matters is the sum of impediments over the path, regardless
    of where in the path they turn up. So long as the potholes aren't
    missed it doesn't matter exactly where they are. For something
    regular, like Belgian block pavement, impediments line up and certain
    paths might find or miss more or less of them, but over the course of
    the path _most_ paths have essentially the same sum of deflections.

    Gravel is a special case, because some of the losses occur in the
    road surface. For that problem a tire that minimizes deflection of
    the road is best.

    Thanks for writing,

    bob prohaska

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Frank Krygowski on Fri Jan 3 08:13:49 2025
    On 1/2/2025 10:36 PM, Frank Krygowski wrote:
    On 1/2/2025 8:45 PM, bp@www.zefox.net wrote:
    Frank Krygowski <frkrygow@sbcglobal.net> wrote:

    One further thought: If we accept (as I do) that jiggling
    the human
    pedaler does cause loss in energy and speed, why aren't
    we all using
    saddles with some sort of damped springing?

    I know suspension seatposts exist, but even those are not
    popular on
    road bikes.

    I'm using a suspension seatpost now, removed from a town
    bike. It's
    slightly more comfortable. No idea if it's more efficient.
    Certainly
    heavier, probably lossy unless I balance pedal effort to
    keep pressure
    on the saddle constant. That difference is small at most.

    FWIW, when coasting - especially on rough downhills - my
    habit is to take some of my weight off the saddle, hoping
    the "suspension" offered by my legs causes less jiggling of
    my body mass, so less energy loss.



    +1 and very relevant to this discussion.

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Radey Shouman@21:1/5 to Frank Krygowski on Fri Jan 3 10:46:33 2025
    Frank Krygowski <frkrygow@sbcglobal.net> writes:

    On 1/2/2025 8:45 PM, bp@www.zefox.net wrote:
    Frank Krygowski <frkrygow@sbcglobal.net> wrote:

    One further thought: If we accept (as I do) that jiggling the human
    pedaler does cause loss in energy and speed, why aren't we all using
    saddles with some sort of damped springing?

    I know suspension seatposts exist, but even those are not popular on
    road bikes.
    I'm using a suspension seatpost now, removed from a town bike. It's
    slightly more comfortable. No idea if it's more efficient. Certainly
    heavier, probably lossy unless I balance pedal effort to keep pressure
    on the saddle constant. That difference is small at most.

    FWIW, when coasting - especially on rough downhills - my habit is to
    take some of my weight off the saddle, hoping the "suspension" offered
    by my legs causes less jiggling of my body mass, so less energy loss.

    To get back to the question of how this might be modeled, it's really complicated. Sometimes all your weight is probably off the saddle,
    meaning that any computation would have to figure out when there was
    contact, and the forces generated by that contact.

    You can read a whole book about it online, if you're ambitious:

    https://www.yastrebov.fr/LECTURES/Yastrebov_NMCM_Wiley_ISTE.pdf

    Not to mention that there is a poorly understood nonlinear control
    mechanism involved, somewhat different for each individual.

    --

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Zen Cycle@21:1/5 to AMuzi on Fri Jan 3 10:41:05 2025
    On 1/2/2025 11:06 AM, AMuzi wrote:
    On 1/2/2025 9:42 AM, bp@www.zefox.net wrote:
    AMuzi <am@yellowjersey.org> wrote:

    I don't have a coherent argument either way but a rumble
    strip test introduces a repeatable experience so that
    various data may be compared.  Each rider on a dirt or
    gravel path, and each ride experience by any given rider, is
    an unique set of impedimenta such that data cannot be as
    readily compared.

    AMuzi <am@yellowjersey.org> wrote:

    I don't have a coherent argument either way but a rumble
    strip test introduces a repeatable experience so that
    various data may be compared.  Each rider on a dirt or
    gravel path, and each ride experience by any given rider, is
    an unique set of impedimenta such that data cannot be as
    readily compared.

    But one can observe that in the case of smooth pavement,
    suspension losses vanish, while hysteresis losses persist.

    In the end a bike is an overdamped resonator excited by the
    pavement and damped by hysteresis, separately in the tire and
    suspenesion. In that limit, suspension would be faster if used
    with very hard tires on very smooth surfaces. In the limit of
    hard tires and no suspension, the dissipative element becomes
    the rider whose elastic properties are apt to be poor, perhaps
    accounting for the apparent slowness of solid tires.

    Use of a rumble strip for testing is equivalent to selecting
    a particular excitation spectrum. Choice of spectrum will affect
    dissipation depending on internal resonances of the bike/rider
    system. A real road likely corresponds to a 1/f spectrum, but
    a rumble strip will likely be something else. How much difference
    that makes isn't clear but it could be estimated using a mechanical
    analogy equivalent circuit of the kind used to model loudspeakers.

    This is a great analysis and reveals a highly problematic aspect of the
    "rumble strip" test. As Bob notes, it's essentially limiting the noise
    input into the system to a somewhat narrow spectral component (though
    the 1/f assumption for real-world is way to broad)

    The idea of using the rumble strip test seems adequate at first, but is
    prone to misleading results. Since the rumble strip sets up a regular
    frequency component, there's a possibility that a resonance or
    cancellation effect can occur which can dramatically skew the results.

    In the world of environmental testing, physical vibration analysis is
    typically broken up into three different stimuli - swept frequency,
    noise*, and environmental specific (usually a combination of noise with
    higher energy components around certain frequencies).

    It's nearly impossible to simulate all the possible real-world
    conditions, which is why the testing regimen includes a sweep - the
    intent being to see any resonances. I've personally witnessed an
    electronic assembly quite nearly disintegrate with the right frequency
    and energy input. The task then was to redesign the piece such that the resonance was damped.

    It's easy to see how this can translate to the rumble strip test. Under
    the right conditions, one might actually see a speed _increase_ as a
    result of a sympathetic resonance.


    A pair of series RLC circuits (one for the road-tire interface
    and a second for the suspension-rider interface) would be a good
    start. I'm not skilled enough to do the calculations, but others
    on this group likely are. The hardest part is apt to be finding
    an equivalent circuit for the rider, who isn't a rigid mass but
    rather a dissipative blob....8-)

    In the old days, we had to do reiterative tests on massive vibration
    tables. These days, FEA software removes the vast amount of guesswork.
    The last few times I've had to conduct these tests I only had to do one
    test twice, and the problem turned out to be an assembly specification
    error rather than inherent design.


    Thanks for reading,

    bob prohaska


    Clever.

    I take from that, you think the actual impact/height change/velocity
    change etc from various irregular surfaces can be quantified for any
    given random gravel (or road) experience and used to compare efficiency
    for other iterations.

    "Real-world" would simulate a more stochastic environment with larger
    "impact" events rather than a regular "sinusoidal" spectrum like a
    rumble strip. Currently, for example, we use this for our truck-mounted electronics:

    https://cvgstrategy.com/wp-content/uploads/2019/08/MIL-STD-810H-Method-514.8-Vibration.pdf

    Refer to page 514.8C-5 (Page 58 in the PDF).


    I hadn't thought of that, but if that's true then the rumble strip test
    isn't necessary for comparison. Which assumes sensors have adequate sensitivity across whatever range and that software for that data truly derives actual impedimenta values.

    Even low-cost accelerometers have incredible accuracy, sensitivity, and repeatability across spectrum they're designed to operate these days. We
    have two 3-axis units accurate to .01G that we paid like $25 each for -
    coupled to a mid-range oscilloscope they give more than adequate results
    for our "warm fuzzy" testing before we send of to a testing lab for 3rd
    party analysis.


    *"Noise" being a broad term meaning quasi-random frequency and amplitude components within limits.

    --
    Add xx to reply

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Zen Cycle on Fri Jan 3 10:17:59 2025
    On 1/3/2025 9:41 AM, Zen Cycle wrote:
    On 1/2/2025 11:06 AM, AMuzi wrote:
    On 1/2/2025 9:42 AM, bp@www.zefox.net wrote:
    AMuzi <am@yellowjersey.org> wrote:

    I don't have a coherent argument either way but a rumble
    strip test introduces a repeatable experience so that
    various data may be compared.  Each rider on a dirt or
    gravel path, and each ride experience by any given
    rider, is
    an unique set of impedimenta such that data cannot be as
    readily compared.

    AMuzi <am@yellowjersey.org> wrote:

    I don't have a coherent argument either way but a rumble
    strip test introduces a repeatable experience so that
    various data may be compared.  Each rider on a dirt or
    gravel path, and each ride experience by any given
    rider, is
    an unique set of impedimenta such that data cannot be as
    readily compared.

    But one can observe that in the case of smooth pavement,
    suspension losses vanish, while hysteresis losses persist.

    In the end a bike is an overdamped resonator excited by the
    pavement and damped by hysteresis, separately in the tire
    and
    suspenesion. In that limit, suspension would be faster if
    used
    with very hard tires on very smooth surfaces. In the
    limit of
    hard tires and no suspension, the dissipative element
    becomes
    the rider whose elastic properties are apt to be poor,
    perhaps
    accounting for the apparent slowness of solid tires.

    Use of a rumble strip for testing is equivalent to selecting
    a particular excitation spectrum. Choice of spectrum will
    affect
    dissipation depending on internal resonances of the bike/
    rider
    system. A real road likely corresponds to a 1/f spectrum,
    but
    a rumble strip will likely be something else. How much
    difference
    that makes isn't clear but it could be estimated using a
    mechanical
    analogy equivalent circuit of the kind used to model
    loudspeakers.

    This is a great analysis and reveals a highly problematic
    aspect of the "rumble strip" test. As Bob notes, it's
    essentially limiting the noise input into the system to a
    somewhat narrow spectral component (though the 1/f
    assumption for real-world is way to broad)

    The idea of using the rumble strip test seems adequate at
    first, but is prone to misleading results. Since the rumble
    strip sets up a regular frequency component, there's a
    possibility that a resonance or cancellation effect can
    occur which can dramatically skew the results.

    In the world of environmental testing, physical vibration
    analysis is typically broken up into three different stimuli
    - swept frequency, noise*, and environmental specific
    (usually a combination of noise with higher energy
    components around certain frequencies).

    It's nearly impossible to simulate all the possible real-
    world conditions, which is why the testing regimen includes
    a sweep - the intent being to see any resonances. I've
    personally witnessed an electronic assembly quite nearly
    disintegrate with the right frequency and energy input. The
    task then was to redesign the piece such that the resonance
    was damped.

    It's easy to see how this can translate to the rumble strip
    test. Under the right conditions, one might actually see a
    speed _increase_ as a result of a sympathetic resonance.


    A pair of series RLC circuits (one for the road-tire
    interface
    and a second for the suspension-rider interface) would be
    a good
    start. I'm not skilled enough to do the calculations, but
    others
    on this group likely are. The hardest part is apt to be
    finding
    an equivalent circuit for the rider, who isn't a rigid
    mass but
    rather a dissipative blob....8-)

    In the old days, we had to do reiterative tests on massive
    vibration tables. These days, FEA software removes the vast
    amount of guesswork. The last few times I've had to conduct
    these tests I only had to do one test twice, and the problem
    turned out to be an assembly specification error rather than
    inherent design.


    Thanks for reading,

    bob prohaska


    Clever.

    I take from that, you think the actual impact/height
    change/velocity change etc from various irregular surfaces
    can be quantified for any given random gravel (or road)
    experience and used to compare efficiency for other
    iterations.

    "Real-world" would simulate a more stochastic environment
    with larger "impact" events rather than a regular
    "sinusoidal" spectrum like a rumble strip. Currently, for
    example, we use this for our truck-mounted electronics:

    https://cvgstrategy.com/wp-content/uploads/2019/08/MIL- STD-810H-Method-514.8-Vibration.pdf

    Refer to page 514.8C-5 (Page 58 in the PDF).


    I hadn't thought of that, but if that's true then the
    rumble strip test isn't necessary for comparison. Which
    assumes sensors have adequate sensitivity across whatever
    range and that software for that data truly derives actual
    impedimenta values.

    Even low-cost accelerometers have incredible accuracy,
    sensitivity, and repeatability across spectrum they're
    designed to operate these days. We have two 3-axis units
    accurate to .01G that we paid like $25 each for - coupled to
    a mid-range oscilloscope they give more than adequate
    results for our "warm fuzzy" testing before we send of to a
    testing lab for 3rd party analysis.


    *"Noise" being a broad term meaning quasi-random frequency
    and amplitude components within limits.


    Thanks I knew nothing about this before our discussion.

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Roger Merriman@21:1/5 to AMuzi on Fri Jan 3 16:38:26 2025
    AMuzi <am@yellowjersey.org> wrote:
    On 1/2/2025 5:45 AM, Roger Merriman wrote:
    Frank Krygowski <frkrygow@sbcglobal.net> wrote:
    Good article from Jan Heine on benefits of wider, softer tires for
    absorbing vibration and lessening suspension losses:

    https://www.renehersecycles.com/the-missing-link-suspension-losses/

    At the time the rumble strip test was published, I expressed some
    skepticism because its roughness is fundamentally different than the
    random roughness of either a rough road or a gravel road. In particular, >>> the rumble strip is all "negative," cut into the smooth surface, while
    rough or gravel roads have both "negative" holes plus "positive" patches >>> or rocks that protrude above the surface. One practical difference is
    that when dealing with only "negative" roughness, higher speeds reduce
    losses. The opposite is true with "positive" roughness.

    But I suppose for demonstrating the fundamental effect, the consistency
    of the rumbles is useful. And the measurements seem valid as long as the >>> test speed is also consistent.

    Not that convinced to be honest, for a starters folks aren’t going to be >> riding rumble strips but by mistake!

    And if you’re going to be real world testing, testing on dirt roads with >> all of the inconsistencies that brings is what gravel riders do. With the
    dips as well as the bumps, plus ruts etc.

    Rumble strip testing seems somewhat misleading ie it’s not that controlled >> nor what riders do.

    As ever claims that they influence pro athletes etc and started the wider
    tire use, IMO it along with disks was adapted by consumers/commuters with
    pro racers lagging behind with adoption and haven’t gone quite as wide, ie >> stopped at 28 for the Pros where as 30/32 are fairly common among club
    riders.


    BTW, Jobst Brandt is mentioned in the article. I recall that in
    discussing rolling resistance here, he insisted that "rolling
    resistance" should be defined _only_ as the losses generated by tire
    rubber's hysteresis. I disagreed, because that implied that solid rubber >>> tires a la 1880, or near infinite tire pressure, or even metal rims with >>> no tire, would be best. Anyone who has ridden an antique solid tire
    "safety" bike knows how slow those tires were.


    As ever is a what you want as well, on the old school road bike, I commute >> on, 28mm felt on the twitchy side 32mm much more planted, the speed
    difference I’m less concerned about, though at that level maybe wider is >> faster? What is faster would depend on road/bike/rider.

    Roger Merriman



    I don't have a coherent argument either way but a rumble
    strip test introduces a repeatable experience so that
    various data may be compared. Each rider on a dirt or
    gravel path, and each ride experience by any given rider, is
    an unique set of impedimenta such that data cannot be as
    readily compared.

    I’d say that’s a fairer test as it’s what riders do, and if the experiment
    shows no difference, that’s fair enough.

    I ride my MTB and Gravel bike on same trails, depending on trail conditions
    and how technical the trail is, will depend on which bike is faster and my times generally group fairly close.

    In short you make an interesting point but it's not
    measurable for comparative purposes.

    Roger Merriman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Catrike Rider@21:1/5 to frkrygow@sbcglobal.net on Fri Jan 3 12:43:25 2025
    On Fri, 3 Jan 2025 11:44:17 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/3/2025 10:46 AM, Radey Shouman wrote:
    Frank Krygowski <frkrygow@sbcglobal.net> writes:

    On 1/2/2025 8:45 PM, bp@www.zefox.net wrote:
    Frank Krygowski <frkrygow@sbcglobal.net> wrote:

    One further thought: If we accept (as I do) that jiggling the human
    pedaler does cause loss in energy and speed, why aren't we all using >>>>> saddles with some sort of damped springing?

    I know suspension seatposts exist, but even those are not popular on >>>>> road bikes.
    I'm using a suspension seatpost now, removed from a town bike. It's
    slightly more comfortable. No idea if it's more efficient. Certainly
    heavier, probably lossy unless I balance pedal effort to keep pressure >>>> on the saddle constant. That difference is small at most.

    FWIW, when coasting - especially on rough downhills - my habit is to
    take some of my weight off the saddle, hoping the "suspension" offered
    by my legs causes less jiggling of my body mass, so less energy loss.

    To get back to the question of how this might be modeled, it's really
    complicated. Sometimes all your weight is probably off the saddle,
    meaning that any computation would have to figure out when there was
    contact, and the forces generated by that contact.

    You can read a whole book about it online, if you're ambitious:

    https://www.yastrebov.fr/LECTURES/Yastrebov_NMCM_Wiley_ISTE.pdf

    Not to mention that there is a poorly understood nonlinear control
    mechanism involved, somewhat different for each individual.

    Wow. Yes, it's probably good to remind ourselves that whatever topic we
    tyros discuss here has probably been the life's work of some true expert.

    <LOL> "True experts" are a dime dozen.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Roger Merriman@21:1/5 to Frank Krygowski on Fri Jan 3 18:17:54 2025
    Frank Krygowski <frkrygow@sbcglobal.net> wrote:
    On 1/3/2025 11:12 AM, cyclintom wrote:

    Coming out of Niles Canyon, you have to ride at around 20 mph Because of
    traffic I was forced to cross a rumble strip with my 28 mm tires and
    came damned close to losing control but it did allow me to let 5 cars
    moving at 45 mph + get past before a constriction. While you're talking
    about taking the lane why don't you come here and try taking the lane?
    You would soon discover, if you're lucky, from a hospital bed that
    California deivers don't like your ideas.

    Ah. We haven't had a "Bicycling is really dangerous _HERE_!" post in
    quite a while.

    So you judged that nearly losing control in front of a 45 mph car was
    safer than legally taking the lane? Yes, my choice would have been
    different, and I've made that choice in <gasp!> California; but
    admittedly not in your super-dangerous neighborhood. When I do that, motorists wait until its safe to pass. Exceptions are vanishingly rare.

    I’d assume most folks would ie use the road than ride in the gutter on the rumble strips! Though I can’t see much evidence of any bar a central line,
    so as ever not sure how/why Tom would be riding in those.

    Seems on a very brief search that some gutters have been used by some
    cyclists as painted bike lanes, which isn’t a wildly good idea at best of times! And are unhappy at the possibility of encountering rumble strips,
    which seems likely to be a poor road all around!

    Do have some painted gutters though Heathrow which i suspect the might
    trick the unwary into thinking they are bike lanes, though it’s a fairly
    car centric type of roads so probably somewhat self selecting, ie I’ve only ever seen folks like myself ie brave folks on road bikes, though it’s a
    very rarely go though on the commute MTB which the gutters are less of no
    no as it’s plush tires are unfazed by drain covered, and one is traveling quite a lot slower, though even so it’s not a terribly wise idea.

    Roger Merriman

    As I often ask, what do you do when riding in a ten foot lane with no shoulder, when an 8.5 foot wide truck approaches from behind? Do you
    jump off the bike and humbly bow?



    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Zen Cycle@21:1/5 to Frank Krygowski on Fri Jan 3 14:04:00 2025
    On 1/3/2025 11:50 AM, Frank Krygowski wrote:
    On 1/3/2025 10:41 AM, Zen Cycle wrote:
    On 1/2/2025 11:06 AM, AMuzi wrote:
    On 1/2/2025 9:42 AM, bp@www.zefox.net wrote:
    AMuzi <am@yellowjersey.org> wrote:

    I don't have a coherent argument either way but a rumble
    strip test introduces a repeatable experience so that
    various data may be compared.  Each rider on a dirt or
    gravel path, and each ride experience by any given rider, is
    an unique set of impedimenta such that data cannot be as
    readily compared.

    AMuzi <am@yellowjersey.org> wrote:

    I don't have a coherent argument either way but a rumble
    strip test introduces a repeatable experience so that
    various data may be compared.  Each rider on a dirt or
    gravel path, and each ride experience by any given rider, is
    an unique set of impedimenta such that data cannot be as
    readily compared.

    But one can observe that in the case of smooth pavement,
    suspension losses vanish, while hysteresis losses persist.

    In the end a bike is an overdamped resonator excited by the
    pavement and damped by hysteresis, separately in the tire and
    suspenesion. In that limit, suspension would be faster if used
    with very hard tires on very smooth surfaces. In the limit of
    hard tires and no suspension, the dissipative element becomes
    the rider whose elastic properties are apt to be poor, perhaps
    accounting for the apparent slowness of solid tires.

    Use of a rumble strip for testing is equivalent to selecting
    a particular excitation spectrum. Choice of spectrum will affect
    dissipation depending on internal resonances of the bike/rider
    system. A real road likely corresponds to a 1/f spectrum, but
    a rumble strip will likely be something else. How much difference
    that makes isn't clear but it could be estimated using a mechanical
    analogy equivalent circuit of the kind used to model loudspeakers.

    This is a great analysis and reveals a highly problematic aspect of
    the "rumble strip" test. As Bob notes, it's essentially limiting the
    noise input into the system to a somewhat narrow spectral component
    (though the 1/f assumption for real-world is way to broad)

    The idea of using the rumble strip test seems adequate at first, but
    is prone to misleading results. Since the rumble strip sets up a
    regular frequency component, there's a possibility that a resonance or
    cancellation effect can occur which can dramatically skew the results.

    In the world of environmental testing, physical vibration analysis is
    typically broken up into three different stimuli - swept frequency,
    noise*, and environmental specific (usually a combination of noise
    with higher energy components around certain frequencies).

    It's nearly impossible to simulate all the possible real-world
    conditions, which is why the testing regimen includes a sweep - the
    intent being to see any resonances.

    Interesting. I suppose the rumble strip test could do a "sweep" by
    riding repeatedly at a wide range of speeds.


    I was thinking we could mount a bike frame with a dent in the top tube
    on an industrial vibration table and see what frequency profile is
    require to pop the dent out....

    https://www.youtube.com/watch?v=3DYvdFp4umM

    But enough serious science, how about fun with resonance?!?!

    https://www.youtube.com/shorts/AJuUIyc73dk

    --
    Add xx to reply

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Zen Cycle@21:1/5 to AMuzi on Fri Jan 3 13:49:46 2025
    On 1/3/2025 11:17 AM, AMuzi wrote:
    On 1/3/2025 9:41 AM, Zen Cycle wrote:
    On 1/2/2025 11:06 AM, AMuzi wrote:
    On 1/2/2025 9:42 AM, bp@www.zefox.net wrote:
    AMuzi <am@yellowjersey.org> wrote:

    I don't have a coherent argument either way but a rumble
    strip test introduces a repeatable experience so that
    various data may be compared.  Each rider on a dirt or
    gravel path, and each ride experience by any given rider, is
    an unique set of impedimenta such that data cannot be as
    readily compared.

    AMuzi <am@yellowjersey.org> wrote:

    I don't have a coherent argument either way but a rumble
    strip test introduces a repeatable experience so that
    various data may be compared.  Each rider on a dirt or
    gravel path, and each ride experience by any given rider, is
    an unique set of impedimenta such that data cannot be as
    readily compared.

    But one can observe that in the case of smooth pavement,
    suspension losses vanish, while hysteresis losses persist.

    In the end a bike is an overdamped resonator excited by the
    pavement and damped by hysteresis, separately in the tire and
    suspenesion. In that limit, suspension would be faster if used
    with very hard tires on very smooth surfaces. In the limit of
    hard tires and no suspension, the dissipative element becomes
    the rider whose elastic properties are apt to be poor, perhaps
    accounting for the apparent slowness of solid tires.

    Use of a rumble strip for testing is equivalent to selecting
    a particular excitation spectrum. Choice of spectrum will affect
    dissipation depending on internal resonances of the bike/ rider
    system. A real road likely corresponds to a 1/f spectrum, but
    a rumble strip will likely be something else. How much difference
    that makes isn't clear but it could be estimated using a mechanical
    analogy equivalent circuit of the kind used to model loudspeakers.

    This is a great analysis and reveals a highly problematic aspect of
    the "rumble strip" test. As Bob notes, it's essentially limiting the
    noise input into the system to a somewhat narrow spectral component
    (though the 1/f assumption for real-world is way to broad)

    The idea of using the rumble strip test seems adequate at first, but
    is prone to misleading results. Since the rumble strip sets up a
    regular frequency component, there's a possibility that a resonance or
    cancellation effect can occur which can dramatically skew the results.

    In the world of environmental testing, physical vibration analysis is
    typically broken up into three different stimuli - swept frequency,
    noise*, and environmental specific (usually a combination of noise
    with higher energy components around certain frequencies).

    It's nearly impossible to simulate all the possible real- world
    conditions, which is why the testing regimen includes a sweep - the
    intent being to see any resonances. I've personally witnessed an
    electronic assembly quite nearly disintegrate with the right frequency
    and energy input. The task then was to redesign the piece such that
    the resonance was damped.

    It's easy to see how this can translate to the rumble strip test.
    Under the right conditions, one might actually see a speed _increase_
    as a result of a sympathetic resonance.


    A pair of series RLC circuits (one for the road-tire interface
    and a second for the suspension-rider interface) would be a good
    start. I'm not skilled enough to do the calculations, but others
    on this group likely are. The hardest part is apt to be finding
    an equivalent circuit for the rider, who isn't a rigid mass but
    rather a dissipative blob....8-)

    In the old days, we had to do reiterative tests on massive vibration
    tables. These days, FEA software removes the vast amount of guesswork.
    The last few times I've had to conduct these tests I only had to do
    one test twice, and the problem turned out to be an assembly
    specification error rather than inherent design.


    Thanks for reading,

    bob prohaska


    Clever.

    I take from that, you think the actual impact/height change/velocity
    change etc from various irregular surfaces can be quantified for any
    given random gravel (or road) experience and used to compare
    efficiency for other iterations.

    "Real-world" would simulate a more stochastic environment with larger
    "impact" events rather than a regular "sinusoidal" spectrum like a
    rumble strip. Currently, for example, we use this for our truck-
    mounted electronics:

    https://cvgstrategy.com/wp-content/uploads/2019/08/MIL- STD-810H-
    Method-514.8-Vibration.pdf

    Refer to page 514.8C-5 (Page 58 in the PDF).


    I hadn't thought of that, but if that's true then the rumble strip
    test isn't necessary for comparison. Which assumes sensors have
    adequate sensitivity across whatever range and that software for that
    data truly derives actual impedimenta values.

    Even low-cost accelerometers have incredible accuracy, sensitivity,
    and repeatability across spectrum they're designed to operate these
    days. We have two 3-axis units accurate to .01G that we paid like $25
    each for - coupled to a mid-range oscilloscope they give more than
    adequate results for our "warm fuzzy" testing before we send of to a
    testing lab for 3rd party analysis.


    *"Noise" being a broad term meaning quasi-random frequency and
    amplitude components within limits.


    Thanks I knew nothing about this before our discussion.


    The lab we go to has a system similar to this:

    https://www.youtube.com/watch?v=AI6svg4lTMo

    --
    Add xx to reply

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From bp@www.zefox.net@21:1/5 to Zen Cycle on Fri Jan 3 19:23:10 2025
    Zen Cycle <funkmaster@hotmail.com> wrote:
    On 1/2/2025 11:06 AM, AMuzi wrote:
    On 1/2/2025 9:42 AM, bp@www.zefox.net wrote:
    AMuzi <am@yellowjersey.org> wrote:

    I don't have a coherent argument either way but a rumble
    strip test introduces a repeatable experience so that
    various data may be compared.  Each rider on a dirt or
    gravel path, and each ride experience by any given rider, is
    an unique set of impedimenta such that data cannot be as
    readily compared.

    AMuzi <am@yellowjersey.org> wrote:

    I don't have a coherent argument either way but a rumble
    strip test introduces a repeatable experience so that
    various data may be compared.  Each rider on a dirt or
    gravel path, and each ride experience by any given rider, is
    an unique set of impedimenta such that data cannot be as
    readily compared.

    But one can observe that in the case of smooth pavement,
    suspension losses vanish, while hysteresis losses persist.

    In the end a bike is an overdamped resonator excited by the
    pavement and damped by hysteresis, separately in the tire and
    suspenesion. In that limit, suspension would be faster if used
    with very hard tires on very smooth surfaces. In the limit of
    hard tires and no suspension, the dissipative element becomes
    the rider whose elastic properties are apt to be poor, perhaps
    accounting for the apparent slowness of solid tires.

    Use of a rumble strip for testing is equivalent to selecting
    a particular excitation spectrum. Choice of spectrum will affect
    dissipation depending on internal resonances of the bike/rider
    system. A real road likely corresponds to a 1/f spectrum, but
    a rumble strip will likely be something else. How much difference
    that makes isn't clear but it could be estimated using a mechanical
    analogy equivalent circuit of the kind used to model loudspeakers.

    This is a great analysis and reveals a highly problematic aspect of the "rumble strip" test. As Bob notes, it's essentially limiting the noise
    input into the system to a somewhat narrow spectral component (though
    the 1/f assumption for real-world is way to broad)


    One can put some bounds on the spatial frequencies of interest. Those
    longer than the tire radius likely don't make it past the tire. Likewise,
    those shorter than the contact patch don't either, though they might
    still excite internal flexural losses in the tire. Neither of those
    constraints is exactly true, but true enough for practical purposes.

    Delta function inputs, like hitting a sharp edge, are physically
    relevant (it happens) but not relevant to analysis of efficiency,
    the first rule of efficiency being "don't crash". 8-)

    The idea of using the rumble strip test seems adequate at first, but is
    prone to misleading results. Since the rumble strip sets up a regular frequency component, there's a possibility that a resonance or
    cancellation effect can occur which can dramatically skew the results.


    In a lightly damped system, yes. in an overdamped system I'd put that
    use case in the same category as hitting a curb. More trouble than it's
    worth apart from avoiding it.

    In the world of environmental testing, physical vibration analysis is typically broken up into three different stimuli - swept frequency,
    noise*, and environmental specific (usually a combination of noise with higher energy components around certain frequencies).

    It's nearly impossible to simulate all the possible real-world
    conditions, which is why the testing regimen includes a sweep - the
    intent being to see any resonances. I've personally witnessed an
    electronic assembly quite nearly disintegrate with the right frequency
    and energy input. The task then was to redesign the piece such that the resonance was damped.

    It's easy to see how this can translate to the rumble strip test. Under
    the right conditions, one might actually see a speed _increase_ as a
    result of a sympathetic resonance.


    I'm not sure of that. Energy injected from the strip must be reflected
    back on the rebound to be recovered. Since there are losses at every
    step of the excitation process I think the rolling resistance will
    always be elevated at resonance, though some modes could have lower
    losses than others. Either way, the rider won't like it.

    A pair of series RLC circuits (one for the road-tire interface
    and a second for the suspension-rider interface) would be a good
    start. I'm not skilled enough to do the calculations, but others
    on this group likely are. The hardest part is apt to be finding
    an equivalent circuit for the rider, who isn't a rigid mass but
    rather a dissipative blob....8-)

    In the old days, we had to do reiterative tests on massive vibration
    tables. These days, FEA software removes the vast amount of guesswork.
    The last few times I've had to conduct these tests I only had to do one
    test twice, and the problem turned out to be an assembly specification
    error rather than inherent design.


    Thanks for reading,

    bob prohaska


    Clever.

    I take from that, you think the actual impact/height change/velocity
    change etc from various irregular surfaces can be quantified for any
    given random gravel (or road) experience and used to compare efficiency
    for other iterations.

    "Real-world" would simulate a more stochastic environment with larger "impact" events rather than a regular "sinusoidal" spectrum like a
    rumble strip. Currently, for example, we use this for our truck-mounted electronics:

    https://cvgstrategy.com/wp-content/uploads/2019/08/MIL-STD-810H-Method-514.8-Vibration.pdf

    Refer to page 514.8C-5 (Page 58 in the PDF).


    I hadn't thought of that, but if that's true then the rumble strip test
    isn't necessary for comparison. Which assumes sensors have adequate
    sensitivity across whatever range and that software for that data truly
    derives actual impedimenta values.

    Even low-cost accelerometers have incredible accuracy, sensitivity, and repeatability across spectrum they're designed to operate these days. We
    have two 3-axis units accurate to .01G that we paid like $25 each for - coupled to a mid-range oscilloscope they give more than adequate results
    for our "warm fuzzy" testing before we send of to a testing lab for 3rd
    party analysis.


    The notion of mounting accelerometers on axles, seatpost and rider is a
    good one if somebody is motivated to do it. It might shed light on the importance of frame stiffness as well.


    *"Noise" being a broad term meaning quasi-random frequency and amplitude components within limits.


    Thanks for reading!

    bob prohaska

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Radey Shouman@21:1/5 to Frank Krygowski on Fri Jan 3 16:16:49 2025
    Frank Krygowski <frkrygow@sbcglobal.net> writes:

    On 1/3/2025 10:41 AM, Zen Cycle wrote:
    On 1/2/2025 11:06 AM, AMuzi wrote:
    On 1/2/2025 9:42 AM, bp@www.zefox.net wrote:
    AMuzi <am@yellowjersey.org> wrote:

    I don't have a coherent argument either way but a rumble
    strip test introduces a repeatable experience so that
    various data may be compared.  Each rider on a dirt or
    gravel path, and each ride experience by any given rider, is
    an unique set of impedimenta such that data cannot be as
    readily compared.

    AMuzi <am@yellowjersey.org> wrote:

    I don't have a coherent argument either way but a rumble
    strip test introduces a repeatable experience so that
    various data may be compared.  Each rider on a dirt or
    gravel path, and each ride experience by any given rider, is
    an unique set of impedimenta such that data cannot be as
    readily compared.

    But one can observe that in the case of smooth pavement,
    suspension losses vanish, while hysteresis losses persist.

    In the end a bike is an overdamped resonator excited by the
    pavement and damped by hysteresis, separately in the tire and
    suspenesion. In that limit, suspension would be faster if used
    with very hard tires on very smooth surfaces. In the limit of
    hard tires and no suspension, the dissipative element becomes
    the rider whose elastic properties are apt to be poor, perhaps
    accounting for the apparent slowness of solid tires.

    Use of a rumble strip for testing is equivalent to selecting
    a particular excitation spectrum. Choice of spectrum will affect
    dissipation depending on internal resonances of the bike/rider
    system. A real road likely corresponds to a 1/f spectrum, but
    a rumble strip will likely be something else. How much difference
    that makes isn't clear but it could be estimated using a mechanical
    analogy equivalent circuit of the kind used to model loudspeakers.
    This is a great analysis and reveals a highly problematic aspect of
    the "rumble strip" test. As Bob notes, it's essentially limiting the
    noise input into the system to a somewhat narrow spectral component
    (though the 1/f assumption for real-world is way to broad)
    The idea of using the rumble strip test seems adequate at first, but
    is prone to misleading results. Since the rumble strip sets up a
    regular frequency component, there's a possibility that a resonance
    or cancellation effect can occur which can dramatically skew the
    results.
    In the world of environmental testing, physical vibration analysis
    is typically broken up into three different stimuli - swept
    frequency, noise*, and environmental specific (usually a combination
    of noise with higher energy components around certain frequencies).
    It's nearly impossible to simulate all the possible real-world
    conditions, which is why the testing regimen includes a sweep - the
    intent being to see any resonances.

    Interesting. I suppose the rumble strip test could do a "sweep" by
    riding repeatedly at a wide range of speeds.

    Or you could just ride one of these:

    https://www.mentalfloss.com/article/73142/musical-roads-5-places-where-streets-sing
    --

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to AMuzi on Fri Jan 3 20:27:04 2025
    On Thu, 2 Jan 2025 10:06:08 -0600, AMuzi <am@yellowjersey.org> wrote:

    I take from that, you think the actual impact/height
    change/velocity change etc from various irregular surfaces
    can be quantified for any given random gravel (or road)
    experience and used to compare efficiency for other iterations.

    This is close, but not quite what you're asking.

    "Energy Harvesting from Bicycle Vibrations by Means of Tuned
    Piezoelectric Generators"
    <https://www.mdpi.com/2079-9292/9/9/1377> <https://mdpi-res.com/d_attachment/electronics/electronics-09-01377/article_deploy/electronics-09-01377-v2.pdf?version=1598515211>

    On PDF page 14, it proclaims:

    "8. Prediction of Generated Power
    The electrical power harvested by a piezo-harvester is very low (in
    the order of a few mW), so highly-efficient power management units
    (PMU) have to be used for energy conversion. The output voltage of the piezo-harvester is a random signal with a main harmonic component at
    the resonance vibration frequency of the cantilever. On the other
    hand, electronic loads (such as a battery for energy storage and/or a
    portable device which can be mounted on a bicycle) are typically fed
    by DC voltage; therefore, interface electronic circuits between the piezo-harvester and the load are made up by a rectifier (for AC to DC
    voltage conversion), an electrolytic capacitor for voltage leveling
    and energy storage, and DC-DC converter for impedance matching with
    the electronic load resistance."

    Page 17 has a table of generated power at various speeds.

    Of course it's possible to optimize the bicycle design, material
    (tire) selection, road surface profile, etc to produce the most power
    output. Presumably, the generated electric power will be used to
    power an electric and mechanical doping system. The problem is that
    when you're starting with milliwatts, it's a long way to go before
    sufficient power can be harvested to make a difference in a race or on
    a ride.

    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to frkrygow@sbcglobal.net on Fri Jan 3 20:46:32 2025
    On Thu, 2 Jan 2025 12:17:22 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    A pair of series RLC circuits (one for the road-tire interface
    and a second for the suspension-rider interface) would be a good
    start. I'm not skilled enough to do the calculations, but others
    on this group likely are.

    I _may_ have been able to do such calculations 50 years ago, but I'm not >sure. I certainly can't do them now.

    Actually, the analogy between a mechanical system and RLC (resistance, inductance and cazapitance) calculations are fairly simple. For
    example:

    "Mechanical-electrical analogies" <https://en.wikipedia.org/wiki/Mechanical%E2%80%93electrical_analogies>

    "Electrical Analogies of Mechanical Systems" <https://www.tutorialspoint.com/control_systems/control_systems_electrical_analogies_mechanical.htm>

    "RLC circuit: Analogy with mechanical systems." (From Brazil) <https://proceedings.sbmac.org.br/sbmac/article/download/134486/3384/0>

    The hardest part is apt to be finding
    an equivalent circuit for the rider, who isn't a rigid mass but
    rather a dissipative blob....8-)

    If you're going to build a computer simulation, there are cut-n-paste mechanical models of various human bodies available.

    "A mechanical model to determine the influence of masses and mass
    distribution on the impact force during running" <https://pubmed.ncbi.nlm.nih.gov/10653036/>
    "Simple spring-damper-mass models have been widely used to simulate
    human locomotion. However, most previous models have not accounted for
    the effect of non-rigid masses (wobbling masses) on impact forces."

    Ok, a running model is not going to work well on a bicycle. So, look
    around for something that's a better fit. I'll admit that I've never
    done anything like this, but I can see how it might be possible to
    model a wobbling blob on a bicycle.

    Also, modeling is NOT the hardest part of the problem. In my never
    humble opinion, the most difficult part is dealing with the large
    number of significant figured necessary to maintain accuracy. I human
    or bicycle model might work accurate to maybe 1/10th of a watt, while
    the power produced by a road bump powered energy harvesting system
    might be on the order of fractions of a milliwatt. This forces the
    human model to be accurate well beyond reasonable limits.



    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to frkrygow@sbcglobal.net on Sat Jan 4 11:06:29 2025
    On Sat, 4 Jan 2025 13:02:24 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/3/2025 11:46 PM, Jeff Liebermann wrote:
    On Thu, 2 Jan 2025 12:17:22 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    Actually, I remember sitting in a Vibrations class taught by my least >favorite professor on the day he brought in a large, demonstration-sized >Analog Computer (Remember those?) to model something we had been
    calculating. To the amusement of the students, he was not able to get
    the thing to work properly.

    Oh yes. That brings back memories of projects that should have died
    on the drawing board. In early college, I couldn't afford a real
    scientific calculator (HP-35). So, I decided to build an analog
    computer in a briefcase. Two 10 turn Helipots with geared dials,
    discrete analog multiplier, three log amps, a large mirrored meter and
    a +/-15v power supply. I couldn't find any photos. It actually
    worked. However, there was some controversy over allowing me to use
    it for exams instead of a slide rule. I had to demonstrate how it
    functioned to the administration (several times), until they
    recognized that it was actually slower than a slide rule and that
    there were no stored formulas or constants. When I was finally
    allowed to use it in an exam, the instructor sabotaged my efforts
    creating questions that squashed the numbers into the upper end of the
    log scale, which made reading the results on the meter rather
    difficult. I gave up and bought a Ti SR-10, which was a mistake: <http://www.vintagecalculators.com/html/texas_instruments_sr-10.html>
    I eventually obtained an HP-35 just in time to have the administration temporarily ban the use of calculators during exams.

    My idea wasn't to model a human body for computation purposes, although >others are probably interested in doing that. I'm thinking more of
    coming up with a physical device, perhaps for weighting a tire during a >rolling drum test, so the test would more accurately reflect behavior of >tires when ridden by a human pedaling a bicycle. Maybe 75 pounds of raw
    meat?

    Ummm... How about a crash test dummy? <https://www.google.com/search?q=crash+test+dummy&udm=2>

    "Cycling crash test dummy"
    <https://www.youtube.com/watch?v=5QWKWXUIPwA>

    Just one problem. They're expensive. Used dummies start at about
    $100,000 and can easily cost $1 million with sensors and
    instrumentation.

    "How Crash Test Dummies Evolved To Cost $1 Million" <https://www.youtube.com/watch?v=iATklZ715H8>

    Maybe you should just use a reinforced mannequin? <https://www.google.com/search?q=human+body+mannequin&udm=2>

    75 lbs of raw meat might work, but I suspect would smell rather bad.
    Perhaps ballistic gel might be a suitable substitute? <https://www.google.com/search?q=ballistic+gel+dummy&num=10&udm=2>
    My guess(tm) is about $2,500 for the dummy and underlying skeleton to
    help hold it together.

    Last resort would be to ask the experts:
    <https://www.humaneticsgroup.com>


    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Frank Krygowski on Sat Jan 4 12:35:20 2025
    On 1/4/2025 12:02 PM, Frank Krygowski wrote:
    On 1/3/2025 11:46 PM, Jeff Liebermann wrote:
    On Thu, 2 Jan 2025 12:17:22 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    A pair of series RLC circuits (one for the road-tire
    interface
    and a second for the suspension-rider interface) would
    be a good
    start. I'm not skilled enough to do the calculations,
    but others
    on this group likely are.

    I _may_ have been able to do such calculations 50 years
    ago, but I'm not
    sure. I certainly can't do them now.

    Actually, the analogy between a mechanical system and RLC
    (resistance,
    inductance and cazapitance) calculations are fairly
    simple.  For
    example:

    "Mechanical-electrical analogies"
    <https://en.wikipedia.org/wiki/
    Mechanical%E2%80%93electrical_analogies>

    "Electrical Analogies of Mechanical Systems"
    <https://www.tutorialspoint.com/control_systems/
    control_systems_electrical_analogies_mechanical.htm>

    "RLC circuit: Analogy with mechanical systems."  (From
    Brazil)
    <https://proceedings.sbmac.org.br/sbmac/article/
    download/134486/3384/0>

    I'm pretty familiar with the RLC analogy to mechanical
    vibrations.

    Actually, I remember sitting in a Vibrations class taught by
    my least favorite professor on the day he brought in a
    large, demonstration-sized Analog Computer (Remember those?)
    to model something we had been calculating. To the amusement
    of the students, he was not able to get the thing to work
    properly.


    The hardest part is apt to be finding
    an equivalent circuit for the rider, who isn't a rigid
    mass but
    rather a dissipative blob....8-)

    If you're going to build a computer simulation, there are
    cut-n-paste
    mechanical models of various human bodies available.

    "A mechanical model to determine the influence of masses
    and mass
    distribution on the impact force during running"
    <https://pubmed.ncbi.nlm.nih.gov/10653036/>
    "Simple spring-damper-mass models have been widely used to
    simulate
    human locomotion. However, most previous models have not
    accounted for
    the effect of non-rigid masses (wobbling masses) on impact
    forces."

    Ok, a running model is not going to work well on a
    bicycle.  So, look
    around for something that's a better fit.  I'll admit that
    I've never
    done anything like this, but I can see how it might be
    possible to
    model a wobbling blob on a bicycle.

    Also, modeling is NOT the hardest part of the problem.  In
    my never
    humble opinion, the most difficult part is dealing with
    the large
    number of significant figured necessary to maintain
    accuracy.  I human
    or bicycle model might work accurate to maybe 1/10th of a
    watt, while
    the power produced by a road bump powered energy
    harvesting system
    might be on the order of fractions of a milliwatt.  This
    forces the
    human model to be accurate well beyond reasonable limits.

    My idea wasn't to model a human body for computation
    purposes, although others are probably interested in doing
    that. I'm thinking more of coming up with a physical device,
    perhaps for weighting a tire during a rolling drum test, so
    the test would more accurately reflect behavior of tires
    when ridden by a human pedaling a bicycle. Maybe 75 pounds
    of raw meat?



    Analog computer?
    Like the fluid logic plate in an automatic transmission?

    https://www.carid.com/acdelco/gm-original-equipment-automatic-transmission-valve-body.html--

    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to AMuzi on Sat Jan 4 11:36:56 2025
    On Sat, 4 Jan 2025 12:35:20 -0600, AMuzi <am@yellowjersey.org> wrote:

    Analog computer?
    Like the fluid logic plate in an automatic transmission?

    https://www.carid.com/acdelco/gm-original-equipment-automatic-transmission-valve-body.html--

    Nope. The automobile automatic transmission is actually a digital
    fluidic switch. When all the inputs and outputs are either on, off,
    in, out, left, right, up down etc, it's digital. There are fluidic
    analog computers. The key difference is how numbers are stored. In a
    digital computer, numbers are stored as discrete numbers. In an
    analog computer, numbers are stored as a range of values that require interpolation to produce an output. When I asked Google the same
    question, I received a rather wide range of answers: <https://www.google.com/search?q=difference+between+analog+and+digital+computer>
    That was a common point of contention as computers were being
    developed (roughly 1960 thru 1990). I don't think anyone successfully
    produced an answer that covered all types of computing devices.
    Instead of an official answer, everyone just gave up by about 1990.

    User manual on an early analog computer including a few useful
    examples. Try to visualize what those problems might look like on a
    slide rule or today's personal computers: <https://www.analogmuseum.org/english/collection/eai/tr10/> <https://www.analogmuseum.org/library/eai_tr-10.pdf>


    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Jeff Liebermann on Sat Jan 4 13:52:37 2025
    On 1/4/2025 1:36 PM, Jeff Liebermann wrote:
    On Sat, 4 Jan 2025 12:35:20 -0600, AMuzi <am@yellowjersey.org> wrote:

    Analog computer?
    Like the fluid logic plate in an automatic transmission?

    https://www.carid.com/acdelco/gm-original-equipment-automatic-transmission-valve-body.html--

    Nope. The automobile automatic transmission is actually a digital
    fluidic switch. When all the inputs and outputs are either on, off,
    in, out, left, right, up down etc, it's digital. There are fluidic
    analog computers. The key difference is how numbers are stored. In a digital computer, numbers are stored as discrete numbers. In an
    analog computer, numbers are stored as a range of values that require interpolation to produce an output. When I asked Google the same
    question, I received a rather wide range of answers: <https://www.google.com/search?q=difference+between+analog+and+digital+computer>
    That was a common point of contention as computers were being
    developed (roughly 1960 thru 1990). I don't think anyone successfully produced an answer that covered all types of computing devices.
    Instead of an official answer, everyone just gave up by about 1990.

    User manual on an early analog computer including a few useful
    examples. Try to visualize what those problems might look like on a
    slide rule or today's personal computers: <https://www.analogmuseum.org/english/collection/eai/tr10/> <https://www.analogmuseum.org/library/eai_tr-10.pdf>



    Ah, such as a slide rule. Got it, thanks.

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Wolfgang Strobl@21:1/5 to All on Sat Jan 4 21:59:45 2025
    Am Sat, 4 Jan 2025 13:02:24 -0500 schrieb Frank Krygowski <frkrygow@sbcglobal.net>:

    On 1/3/2025 11:46 PM, Jeff Liebermann wrote:
    On Thu, 2 Jan 2025 12:17:22 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    Actually, I remember sitting in a Vibrations class taught by my least >favorite professor on the day he brought in a large, demonstration-sized >Analog Computer (Remember those?) to model something we had been
    calculating. To the amusement of the students, he was not able to get
    the thing to work properly.

    Most probably it wasn't as large and as heavy as this one. :-) <https://www.mystrobl.de/Plone/wolfgang-strobl/fotos/geraete/eai-pacer-231rv/img-2854-img-2856.jpeg>


    --
    Thank you for observing all safety precautions

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Frank Krygowski on Sat Jan 4 15:44:24 2025
    On 1/4/2025 3:28 PM, Frank Krygowski wrote:
    On 1/4/2025 2:52 PM, AMuzi wrote:
    On 1/4/2025 1:36 PM, Jeff Liebermann wrote:
    On Sat, 4 Jan 2025 12:35:20 -0600, AMuzi
    <am@yellowjersey.org> wrote:

    User manual on an early analog computer including a few
    useful
    examples.  Try to visualize what those problems might
    look like on a
    slide rule or today's personal computers:
    <https://www.analogmuseum.org/english/collection/eai/tr10/>
    <https://www.analogmuseum.org/library/eai_tr-10.pdf>

    Ah, such as a slide rule. Got it, thanks.

    The device I was talking about was nothing like a slide
    rule. It looked vaguely like the one in Jeff's last link
    above, but the classroom demonstrator was much larger -
    maybe 3' x 4' IIRC - with much bigger knobs (4" diameter?)
    and meters.

    We were talking about electrical analogies for vibrating
    masses, and that's one of the things the analog computer
    could simulate. One would have to calculate the values of
    voltage, inductance and resistance to correctly simulate the
    damped spring-mass system, set initial conditions, then let
    the circuit run. The system's meters would then swing back
    and forth in a manner analogous to the position of the mass.
    All this was before digital computers were desktop devices.

    (In those days, the programs I wrote for vaguely similar
    problems were room sized and run by full time technicians,
    and I'd turn in a program stored as a thick deck of punched
    cards, hoping output would be ready the next day.)

    As I recall, we students never did any actual work with that
    analog computer.


    Right. The values shown on that meter with a swinging
    needle are analog, not digital; the scale and range can
    approximate the value to an acceptable degree without a
    specific digital value.


    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to frkrygow@sbcglobal.net on Sat Jan 4 15:20:03 2025
    On Sat, 4 Jan 2025 16:28:21 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/4/2025 2:52 PM, AMuzi wrote:
    On 1/4/2025 1:36 PM, Jeff Liebermann wrote:
    On Sat, 4 Jan 2025 12:35:20 -0600, AMuzi <am@yellowjersey.org> wrote:

    User manual on an early analog computer including a few useful
    examples. Try to visualize what those problems might look like on a
    slide rule or today's personal computers:
    <https://www.analogmuseum.org/english/collection/eai/tr10/>
    <https://www.analogmuseum.org/library/eai_tr-10.pdf>

    Ah, such as a slide rule. Got it, thanks.

    The device I was talking about was nothing like a slide rule.

    I called the analog computer that I build in a briefcase an
    "electronic slide rule". I didn't want to, but that made it more
    acceptable to the college bureaucracy.

    It looked
    vaguely like the one in Jeff's last link above, but the classroom >demonstrator was much larger - maybe 3' x 4' IIRC - with much bigger
    knobs (4" diameter?) and meters.

    I couldn't find anything with such huge knobs. Maybe something like
    these from Edmund Scientific? <https://www.google.com/search?q=edmund+scientific+analog+computer&udm=2>

    Would you believe a Heathkit EC-1 analog computer? <https://www.google.com/search?q=heathkit+ec-1&udm=2> <https://www.nutsvolts.com/magazine/article/may2016_heathkit_restoration>
    Fig 7 is a bouncing ball simulation, which is similar to the bouncing
    bicycle simulation.

    We were talking about electrical analogies for vibrating masses, and
    that's one of the things the analog computer could simulate. One would
    have to calculate the values of voltage, inductance and resistance to >correctly simulate the damped spring-mass system, set initial
    conditions, then let the circuit run. The system's meters would then
    swing back and forth in a manner analogous to the position of the mass.
    All this was before digital computers were desktop devices.

    Meters? Too crude. We used an oscilloscope or X-Y pen plotter.

    (In those days, the programs I wrote for vaguely similar problems were
    room sized and run by full time technicians, and I'd turn in a program
    stored as a thick deck of punched cards, hoping output would be ready
    the next day.)

    As I recall, we students never did any actual work with that analog
    computer.

    We did. My guess(tm) that would 1969. We had groups of 5 or 6
    students sharing one machine. I got some extra experience because I
    worked for the "calibration department" repairing them. The problem
    was we had a large number of foreign exchange students from Iraq. Most
    had never done any manual labor or learned to use tools. When faced
    with a knob that had reached its end of rotation, they simply applied
    more force to help it rotate. That usually broke the expensive 10
    turn potentiometer (Helipot).
    <https://www.google.com/search?q=helipot&udm=2>

    I was the idiot who found a solution to the broken potentiometer
    problem. Between the knob/turn_counter and the pot was a short shaft extension. I machined a few of these and added a plastic shear pin.
    If they hit the end of rotation and break the pin, all they had to do
    was rotate everything full counter clockwise, push the pin out of the
    hole, and replace it with a new pin (or toothpick). The reason I was
    an idiot was because I had found the solution, I sentence to working
    overtime retrofitting all the analog computers with shear pins.

    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to All on Sat Jan 4 16:06:26 2025
    On Sat, 04 Jan 2025 11:06:29 -0800, Jeff Liebermann <jeffl@cruzio.com>
    wrote:

    In early college, I couldn't afford a real
    scientific calculator (HP-35).
    (...)
    I gave up and bought a Ti SR-10, which was a mistake: ><http://www.vintagecalculators.com/html/texas_instruments_sr-10.html>
    I eventually obtained an HP-35 just in time to have the administration >temporarily ban the use of calculators during exams.

    Something is wrong with my dates. The HP-35 was introduced in Feb
    1972. I graduated from college in June 1971. I attended a few months
    of teacher prep classes in late 1971, but don't recall having or
    needing an HP-35 calculator during that time. There is no possible
    way I could have owned an HP-35 before it was offered for sale in Feb
    1972.

    It seems that I did something wrong, mangled my dates, faulty memory,
    etc. That's not surprising for something that happened 54 years ago.
    The problem is I don't know where I made the mistake. I still have
    both the SR-10 and HP-35 calculators and boxes. I'll see if there are
    some dates on the paperwork in the boxes. Meanwhile, please ignore my
    comments related to the HP-35.

    Sorry for the confusion.




    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to frkrygow@sbcglobal.net on Sat Jan 4 16:28:44 2025
    On Sat, 4 Jan 2025 16:41:25 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/4/2025 2:06 PM, Jeff Liebermann wrote:
    I gave up and bought a Ti SR-10, which was a mistake:
    <http://www.vintagecalculators.com/html/texas_instruments_sr-10.html>
    I eventually obtained an HP-35 just in time to have the administration
    temporarily ban the use of calculators during exams.

    As a HP fan, my reaction to "... bought a TI SR-10" was "ewww."

    Brain rot again. The Ti SR-10 was introduced in Nov 1972. I
    graduated from college (Cal Poly Pomona) in June 1971. Obviously, I
    couldn't have had that calculator while still in college. Yet, 54
    years later, that's the way I remember it.

    Same as the HP-35. Please ignore my comments about the SR-10 until I
    have some time to untangle the mess. Again, sorry about the
    confusion.

    You may be vaguely interested in this company: >https://www.swissmicros.com/products

    Nice products. I've played with a few of them owned by acquaintances.
    Good display and some useful added features. I wish it had a
    backlighted display. However, they're too expensive for my minimal
    use.

    Replica HP calculators. They seem to have a good reputation.

    I'm not sure if they're a replica, clone, compatible device, or
    work-alike. I haven't read the reviews, but what little I've skimmed,
    it seems like a well built product.

    I'm doing well with my HP48G, thank you. But I still miss my HP11C that
    was stolen while I worked at a little company with bad security.

    I'll look in my collection and see if I have an HP11C. Offhand, I
    don't think so. There are some on eBay. The one's that offer the
    least risk and are likely to work are $75 and up.

    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to All on Sat Jan 4 16:40:33 2025
    On Sat, 04 Jan 2025 16:28:44 -0800, Jeff Liebermann <jeffl@cruzio.com>
    wrote:

    I'll look in my collection and see if I have an HP11C. Offhand, I
    don't think so. There are some on eBay. The one's that offer the
    least risk and are likely to work are $75 and up.

    There's also the PX-11c. I don't know anything about how well it's
    built or how it works.

    <https://www.google.com/search?q=px-11c>
    It's an HP-11c clone sold as a kit. $70 plus tax and shipping. Claims
    it's 4 to 5 times faster than the HP version:

    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From zen cycle@21:1/5 to Jeff Liebermann on Sun Jan 5 05:49:53 2025
    On 1/4/2025 2:36 PM, Jeff Liebermann wrote:
    On Sat, 4 Jan 2025 12:35:20 -0600, AMuzi <am@yellowjersey.org> wrote:

    Analog computer?
    Like the fluid logic plate in an automatic transmission?

    https://www.carid.com/acdelco/gm-original-equipment-automatic-transmission-valve-body.html--

    Nope. The automobile automatic transmission is actually a digital
    fluidic switch. When all the inputs and outputs are either on, off,
    in, out, left, right, up down etc, it's digital.

    Except for CVTs



    There are fluidic
    analog computers. The key difference is how numbers are stored. In a digital computer, numbers are stored as discrete numbers. In an
    analog computer, numbers are stored as a range of values that require interpolation to produce an output. When I asked Google the same
    question, I received a rather wide range of answers: <https://www.google.com/search?q=difference+between+analog+and+digital+computer>
    That was a common point of contention as computers were being
    developed (roughly 1960 thru 1990). I don't think anyone successfully produced an answer that covered all types of computing devices.
    Instead of an official answer, everyone just gave up by about 1990.

    User manual on an early analog computer including a few useful
    examples. Try to visualize what those problems might look like on a
    slide rule or today's personal computers: <https://www.analogmuseum.org/english/collection/eai/tr10/> <https://www.analogmuseum.org/library/eai_tr-10.pdf>



    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From zen cycle@21:1/5 to Jeff Liebermann on Sun Jan 5 05:40:18 2025
    On 1/4/2025 6:20 PM, Jeff Liebermann wrote:
    On Sat, 4 Jan 2025 16:28:21 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/4/2025 2:52 PM, AMuzi wrote:
    On 1/4/2025 1:36 PM, Jeff Liebermann wrote:
    On Sat, 4 Jan 2025 12:35:20 -0600, AMuzi <am@yellowjersey.org> wrote:

    User manual on an early analog computer including a few useful
    examples.  Try to visualize what those problems might look like on a
    slide rule or today's personal computers:
    <https://www.analogmuseum.org/english/collection/eai/tr10/>
    <https://www.analogmuseum.org/library/eai_tr-10.pdf>

    Ah, such as a slide rule. Got it, thanks.

    The device I was talking about was nothing like a slide rule.

    I called the analog computer that I build in a briefcase an
    "electronic slide rule". I didn't want to, but that made it more
    acceptable to the college bureaucracy.

    It looked
    vaguely like the one in Jeff's last link above, but the classroom
    demonstrator was much larger - maybe 3' x 4' IIRC - with much bigger
    knobs (4" diameter?) and meters.

    I couldn't find anything with such huge knobs. Maybe something like
    these from Edmund Scientific? <https://www.google.com/search?q=edmund+scientific+analog+computer&udm=2>

    Would you believe a Heathkit EC-1 analog computer? <https://www.google.com/search?q=heathkit+ec-1&udm=2> <https://www.nutsvolts.com/magazine/article/may2016_heathkit_restoration>
    Fig 7 is a bouncing ball simulation, which is similar to the bouncing
    bicycle simulation.

    We were talking about electrical analogies for vibrating masses, and
    that's one of the things the analog computer could simulate. One would
    have to calculate the values of voltage, inductance and resistance to
    correctly simulate the damped spring-mass system, set initial
    conditions, then let the circuit run. The system's meters would then
    swing back and forth in a manner analogous to the position of the mass.
    All this was before digital computers were desktop devices.

    Meters? Too crude. We used an oscilloscope or X-Y pen plotter.

    (In those days, the programs I wrote for vaguely similar problems were
    room sized and run by full time technicians, and I'd turn in a program
    stored as a thick deck of punched cards, hoping output would be ready
    the next day.)

    As I recall, we students never did any actual work with that analog
    computer.

    We did. My guess(tm) that would 1969. We had groups of 5 or 6
    students sharing one machine. I got some extra experience because I
    worked for the "calibration department" repairing them. The problem
    was we had a large number of foreign exchange students from Iraq. Most
    had never done any manual labor or learned to use tools. When faced
    with a knob that had reached its end of rotation, they simply applied
    more force to help it rotate. That usually broke the expensive 10
    turn potentiometer (Helipot).
    <https://www.google.com/search?q=helipot&udm=2>

    I was the idiot who found a solution to the broken potentiometer
    problem. Between the knob/turn_counter and the pot was a short shaft extension. I machined a few of these and added a plastic shear pin.
    If they hit the end of rotation and break the pin, all they had to do
    was rotate everything full counter clockwise, push the pin out of the
    hole, and replace it with a new pin (or toothpick). The reason I was
    an idiot was because I had found the solution, I sentence to working
    overtime retrofitting all the analog computers with shear pins.


    I worked with a curmudgeonly older engineer many years ago who quipped
    "do a shitty job well and it's yours forever".

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From zen cycle@21:1/5 to Jeff Liebermann on Sun Jan 5 05:35:29 2025
    On 1/3/2025 11:46 PM, Jeff Liebermann wrote:

    In my never humble opinion,

    IMNHO....Nice...The Interwebs newest intialism.....

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Jeff Liebermann on Sun Jan 5 09:57:46 2025
    On 1/4/2025 5:20 PM, Jeff Liebermann wrote:
    On Sat, 4 Jan 2025 16:28:21 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/4/2025 2:52 PM, AMuzi wrote:
    On 1/4/2025 1:36 PM, Jeff Liebermann wrote:
    On Sat, 4 Jan 2025 12:35:20 -0600, AMuzi <am@yellowjersey.org> wrote:

    User manual on an early analog computer including a few useful
    examples.  Try to visualize what those problems might look like on a
    slide rule or today's personal computers:
    <https://www.analogmuseum.org/english/collection/eai/tr10/>
    <https://www.analogmuseum.org/library/eai_tr-10.pdf>

    Ah, such as a slide rule. Got it, thanks.

    The device I was talking about was nothing like a slide rule.

    I called the analog computer that I build in a briefcase an
    "electronic slide rule". I didn't want to, but that made it more
    acceptable to the college bureaucracy.

    It looked
    vaguely like the one in Jeff's last link above, but the classroom
    demonstrator was much larger - maybe 3' x 4' IIRC - with much bigger
    knobs (4" diameter?) and meters.

    I couldn't find anything with such huge knobs. Maybe something like
    these from Edmund Scientific? <https://www.google.com/search?q=edmund+scientific+analog+computer&udm=2>

    Would you believe a Heathkit EC-1 analog computer? <https://www.google.com/search?q=heathkit+ec-1&udm=2> <https://www.nutsvolts.com/magazine/article/may2016_heathkit_restoration>
    Fig 7 is a bouncing ball simulation, which is similar to the bouncing
    bicycle simulation.

    We were talking about electrical analogies for vibrating masses, and
    that's one of the things the analog computer could simulate. One would
    have to calculate the values of voltage, inductance and resistance to
    correctly simulate the damped spring-mass system, set initial
    conditions, then let the circuit run. The system's meters would then
    swing back and forth in a manner analogous to the position of the mass.
    All this was before digital computers were desktop devices.

    Meters? Too crude. We used an oscilloscope or X-Y pen plotter.

    (In those days, the programs I wrote for vaguely similar problems were
    room sized and run by full time technicians, and I'd turn in a program
    stored as a thick deck of punched cards, hoping output would be ready
    the next day.)

    As I recall, we students never did any actual work with that analog
    computer.

    We did. My guess(tm) that would 1969. We had groups of 5 or 6
    students sharing one machine. I got some extra experience because I
    worked for the "calibration department" repairing them. The problem
    was we had a large number of foreign exchange students from Iraq. Most
    had never done any manual labor or learned to use tools. When faced
    with a knob that had reached its end of rotation, they simply applied
    more force to help it rotate. That usually broke the expensive 10
    turn potentiometer (Helipot).
    <https://www.google.com/search?q=helipot&udm=2>

    I was the idiot who found a solution to the broken potentiometer
    problem. Between the knob/turn_counter and the pot was a short shaft extension. I machined a few of these and added a plastic shear pin.
    If they hit the end of rotation and break the pin, all they had to do
    was rotate everything full counter clockwise, push the pin out of the
    hole, and replace it with a new pin (or toothpick). The reason I was
    an idiot was because I had found the solution, I sentence to working
    overtime retrofitting all the analog computers with shear pins.


    Thank you, especially for the Heath Kit page (although much
    of that went beyond my understanding). The reader comments
    were great, especially the last one!

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Frank Krygowski on Sun Jan 5 11:27:48 2025
    On 1/5/2025 10:25 AM, Frank Krygowski wrote:
    On 1/5/2025 5:40 AM, zen cycle wrote:

    I worked with a curmudgeonly older engineer many years ago
    who quipped "do a shitty job well and it's yours forever".

    One of my colleagues had a sign in his classroom:
    "If you don't have time to do it right, when will you have
    time to do it over?"

    Per Ray Gasiorowsky, "There are two ways to do anything. The
    right way and again."

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to funkmasterxx@hotmail.com on Sun Jan 5 12:46:28 2025
    On Sun, 5 Jan 2025 05:40:18 -0500, zen cycle
    <funkmasterxx@hotmail.com> wrote:

    On 1/4/2025 6:20 PM, Jeff Liebermann wrote:
    I was the idiot who found a solution to the broken potentiometer
    problem. Between the knob/turn_counter and the pot was a short shaft
    extension. I machined a few of these and added a plastic shear pin.
    If they hit the end of rotation and break the pin, all they had to do
    was rotate everything full counter clockwise, push the pin out of the
    hole, and replace it with a new pin (or toothpick). The reason I was
    an idiot was because I had found the solution, I sentence to working
    overtime retrofitting all the analog computers with shear pins.

    I worked with a curmudgeonly older engineer many years ago who quipped
    "do a shitty job well and it's yours forever".

    True. However, that wasn't the problem. No sooner had I demonstrated
    my solution, the calibration lab (and repair) manager was accosted by
    literally everyone involved, each of which had a better idea. Since
    everyone was also busy with something else and therefore could not
    spare the time to prototype and test their ideas, and since I had
    somehow managed to find some time to build and test my idea, and since
    the manager didn't want anyone wasting their precious time working on
    the problem, it was logically determined that I should be the chosen
    one to build and test everyone else's ideas. Swell. Just what I
    didn't need. Somehow, the Dean of Engineering became involved
    (because blaming the Iraqi exchange students for creating the problem
    was politically incorrect) and ordered that everyone working on it
    should find something else to do. My good enough solution was
    declared "good enough" and installed on all the Helipots that showed
    damage (i.e. held together with electrical tape) until we ran out of
    parts. I wasn't involved and moved on to (create) bigger and better
    problems.


    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to AMuzi on Sun Jan 5 12:58:28 2025
    On Sun, 5 Jan 2025 09:57:46 -0600, AMuzi <am@yellowjersey.org> wrote:

    On 1/4/2025 5:20 PM, Jeff Liebermann wrote:

    <https://www.nutsvolts.com/magazine/article/may2016_heathkit_restoration>

    Thank you, especially for the Heath Kit page (although much
    of that went beyond my understanding). The reader comments
    were great, especially the last one!

    Bad news. Most sundials are certainly analog, but there are some
    almost digital exceptions: <https://www.etsy.com/listing/976269465/digital-sundial> <https://en.wikipedia.org/wiki/Digital_sundial> <https://www.hineslab.com/optical-projects/digital-sundial-2/> <https://www.reddit.com/r/mildlyinteresting/comments/atmurl/digital_sundial/> More of the same:
    <https://www.google.com/search?q=digital+sundial&udm=2>




    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to AMuzi on Sun Jan 5 13:26:09 2025
    On Sun, 5 Jan 2025 11:27:48 -0600, AMuzi <am@yellowjersey.org> wrote:

    On 1/5/2025 10:25 AM, Frank Krygowski wrote:
    On 1/5/2025 5:40 AM, zen cycle wrote:

    I worked with a curmudgeonly older engineer many years ago
    who quipped "do a shitty job well and it's yours forever".

    One of my colleagues had a sign in his classroom:
    "If you don't have time to do it right, when will you have
    time to do it over?"

    Per Ray Gasiorowsky, "There are two ways to do anything. The
    right way and again."

    One of my former employers was very much into "first to market"
    project design. Everything we did was fast and furious. His tag line
    was something like "Do it right the first time because there won't be
    a second time if the company goes broke".

    This was somewhat put to the test when we were working on the
    AN/SRD-22 Doppler direction finder for the USCG (US Coast Guard): <https://www.learnbydestroying.com/jeffl/AN-SRD-22/>
    We were at the point where the PCB's were being laid out when I had a
    better idea, which would eliminate two of the major PCB's (printed
    circuit boards), replace a control cable full of wires with one coax
    cable, and substantially reduce the complexity and cost. I camped out
    in my office for a weekend and built a working proof of concept
    prototype.

    Everyone agreed that it was better in every way. The problem was that
    if we stopped the design process and switched to my new and improved
    design, we ran the risk of missing the delivery deadline. It took a
    few days to debate the change in direction, but when done, management
    decided to go with the superior product. Amazingly, we met the first
    article delivery deadline without any fatalities. It was about 3(?)
    months between the decision and delivery. I got very little sleep and
    was sweating bullets for the duration.

    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to funkmasterxx@hotmail.com on Sun Jan 5 13:39:49 2025
    On Sun, 5 Jan 2025 05:35:29 -0500, zen cycle
    <funkmasterxx@hotmail.com> wrote:

    On 1/3/2025 11:46 PM, Jeff Liebermann wrote:

    In my never humble opinion,

    IMNHO....Nice...The Interwebs newest intialism.....

    I think you mean initialism. <https://www.merriam-webster.com/dictionary/initialism>
    I've been using "in my never humble opinion" in its fully expanded
    form for probably 50 years. I typically the phrase only once per
    posting, so there's no need for acronymization.


    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From zen cycle@21:1/5 to Jeff Liebermann on Mon Jan 6 04:57:32 2025
    On 1/5/2025 4:39 PM, Jeff Liebermann wrote:
    On Sun, 5 Jan 2025 05:35:29 -0500, zen cycle
    <funkmasterxx@hotmail.com> wrote:

    On 1/3/2025 11:46 PM, Jeff Liebermann wrote:

    In my never humble opinion,

    IMNHO....Nice...The Interwebs newest intialism.....

    I think you mean initialism.

    Indeed I did. It was a typo.

    <https://www.merriam-webster.com/dictionary/initialism>
    I've been using "in my never humble opinion" in its fully expanded
    form for probably 50 years. I typically the phrase only once per
    posting, so there's no need for acronymization.


    I think you mean _an_ acronymization. :)

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Frank Krygowski on Mon Jan 6 07:40:50 2025
    On 1/2/2025 11:17 AM, Frank Krygowski wrote:
    On 1/2/2025 11:06 AM, AMuzi wrote:
    On 1/2/2025 9:42 AM, bp@www.zefox.net wrote:

    But one can observe that in the case of smooth pavement,
    suspension losses vanish, while hysteresis losses persist.

    I think that would be true only if the smooth pavement were
    as smooth as a linoleum floor. Or a wooden track. IIRC, what
    got Jan Heine started on investigations of rolling
    resistance vs. tire width was coast-down tests on a Soapbox
    Derby track. I suspect that was quite smooth. Soapbox cars
    have hard tires and no suspension, AFAIK.

    In the end a bike is an overdamped resonator excited by the
    pavement and damped by hysteresis, separately in the tire
    and
    suspenesion. In that limit, suspension would be faster if
    used
    with very hard tires on very smooth surfaces. In the
    limit of
    hard tires and no suspension, the dissipative element
    becomes
    the rider whose elastic properties are apt to be poor,
    perhaps
    accounting for the apparent slowness of solid tires.

    Use of a rumble strip for testing is equivalent to selecting
    a particular excitation spectrum. Choice of spectrum will
    affect
    dissipation depending on internal resonances of the bike/
    rider
    system. A real road likely corresponds to a 1/f spectrum,
    but
    a rumble strip will likely be something else. How much
    difference
    that makes isn't clear but it could be estimated using a
    mechanical
    analogy equivalent circuit of the kind used to model
    loudspeakers.

    A pair of series RLC circuits (one for the road-tire
    interface
    and a second for the suspension-rider interface) would be
    a good
    start. I'm not skilled enough to do the calculations, but
    others
    on this group likely are.

    I _may_ have been able to do such calculations 50 years ago,
    but I'm not sure. I certainly can't do them now.

    The hardest part is apt to be finding
    an equivalent circuit for the rider, who isn't a rigid
    mass but
    rather a dissipative blob....8-)

    I actually think physically modeling that dissipative blob
    might be valuable for the tire industry. Using such a blob
    to apply weight during a rolling drum test might give better
    information than what those tests give now.


    Clever.

    I take from that, you think the actual impact/height
    change/velocity change etc from various irregular surfaces
    can be quantified for any given random gravel (or road)
    experience and used to compare efficiency for other
    iterations.

    I hadn't thought of that, but if that's true then the
    rumble strip test isn't necessary for comparison. Which
    assumes sensors have adequate sensitivity across whatever
    range and that software for that data truly derives actual
    impedimenta values.

    There are ways of quantifying roughness, with varying
    scales, varying tools. I'm most familiar with roughness
    measurement of machined parts, with tools varying from
    sample cards for "fingernail" test comparisons, to RMS
    readers akin to phonograph needles or laser scattering devices.

    https://en.wikipedia.org/wiki/Surface_roughness

    ISTR reading about systems for evaluating pavement fairly
    crudely, as in whether it should be repaved or not. I don't
    know of a system actually used for measuring pavement
    roughness at a scale affecting bike tire choice.


    Tangential, but this was just sent to me and I found it
    fascinating, Under 2 minutes:

    https://www.youtube.com/shorts/TGuxwgUyu2A

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to funkmasterxx@hotmail.com on Mon Jan 6 13:19:32 2025
    On Mon, 6 Jan 2025 04:57:32 -0500, zen cycle
    <funkmasterxx@hotmail.com> wrote:

    On 1/5/2025 4:39 PM, Jeff Liebermann wrote:
    On Sun, 5 Jan 2025 05:35:29 -0500, zen cycle
    <funkmasterxx@hotmail.com> wrote:

    On 1/3/2025 11:46 PM, Jeff Liebermann wrote:

    In my never humble opinion,

    IMNHO....Nice...The Interwebs newest intialism.....

    I think you mean initialism.

    Indeed I did. It was a typo.

    <https://www.merriam-webster.com/dictionary/initialism>
    I've been using "in my never humble opinion" in its fully expanded
    form for probably 50 years. I typically the phrase only once per
    posting, so there's no need for acronymization.

    I left out a word. It should read "I typically use the phrase..."
    Thanks for not noticing.

    I think you mean _an_ acronymization. :)

    No. I consider acronymization as the process of converting a human
    readable and easily understood phrase into a cryptic, difficult to
    decode and sometimes clever acronym as "acronymization". <https://en.wiktionary.org/wiki/acronymization>

    Bike Culture Acronyms: <https://www.bikeforums.net/general-cycling-discussion/400164-bike-culture-acronyms.html>


    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to frkrygow@sbcglobal.net on Thu Jan 9 21:20:12 2025
    On Sat, 4 Jan 2025 20:51:16 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/4/2025 7:28 PM, Jeff Liebermann wrote:

    I'll look in my collection and see if I have an HP11C. Offhand, I
    don't think so. There are some on eBay. The one's that offer the
    least risk and are likely to work are $75 and up.

    No need to check. When that calculator got stolen, the guys running the >company bought me a replacement. By then, the 11C was no longer
    marketed, so I ended up with an HP 32S II, one of the few RPN machines
    still in their line, IIRC.

    The 11C seemed bulletproof, but this 32S is a bit flaky. Its the
    calculator I keep in my workshop drawer, so it's not used very often.
    But it seems that at least a third of the time I want to use it, it
    refuses to turn on. I've slipped a little note in its case with notes on
    the recovery procedure.

    When it flakes out, I'll usually just pull out my Android phone and run
    the HP 48G emulator. My main complaint about that one is that it isn't
    really programmable - or at least, it doesn't retain programs when the
    app is killed.

    I finally remembered to look into my boxes of HP calculators. I don't
    have an HP 32SII or something comparable. In the scientific
    calculator section, I found an HP 31E and a 32E. Both of these are
    low end RPN calculators with LED displays. The 31E had a battery leak
    at some time in the past and will require that I do some battery
    contact rebuilding. I can provide a complete list of what I have in
    stock.


    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From zen cycle@21:1/5 to Jeff Liebermann on Fri Jan 10 05:19:53 2025
    On 1/10/2025 12:20 AM, Jeff Liebermann wrote:
    On Sat, 4 Jan 2025 20:51:16 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/4/2025 7:28 PM, Jeff Liebermann wrote:

    I'll look in my collection and see if I have an HP11C. Offhand, I
    don't think so. There are some on eBay. The one's that offer the
    least risk and are likely to work are $75 and up.

    No need to check. When that calculator got stolen, the guys running the
    company bought me a replacement. By then, the 11C was no longer
    marketed, so I ended up with an HP 32S II, one of the few RPN machines
    still in their line, IIRC.

    The 11C seemed bulletproof, but this 32S is a bit flaky. Its the
    calculator I keep in my workshop drawer, so it's not used very often.
    But it seems that at least a third of the time I want to use it, it
    refuses to turn on. I've slipped a little note in its case with notes on
    the recovery procedure.

    When it flakes out, I'll usually just pull out my Android phone and run
    the HP 48G emulator. My main complaint about that one is that it isn't
    really programmable - or at least, it doesn't retain programs when the
    app is killed.

    I finally remembered to look into my boxes of HP calculators. I don't
    have an HP 32SII or something comparable. In the scientific
    calculator section, I found an HP 31E and a 32E. Both of these are
    low end RPN calculators with LED displays. The 31E had a battery leak
    at some time in the past and will require that I do some battery
    contact rebuilding. I can provide a complete list of what I have in
    stock.


    I never could get used to the RPN data entry method

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to funkmasterxx@hotmail.com on Fri Jan 10 07:25:48 2025
    On Fri, 10 Jan 2025 05:19:53 -0500, zen cycle
    <funkmasterxx@hotmail.com> wrote:

    On 1/10/2025 12:20 AM, Jeff Liebermann wrote:
    On Sat, 4 Jan 2025 20:51:16 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/4/2025 7:28 PM, Jeff Liebermann wrote:

    I'll look in my collection and see if I have an HP11C. Offhand, I
    don't think so. There are some on eBay. The one's that offer the
    least risk and are likely to work are $75 and up.

    No need to check. When that calculator got stolen, the guys running the
    company bought me a replacement. By then, the 11C was no longer
    marketed, so I ended up with an HP 32S II, one of the few RPN machines
    still in their line, IIRC.

    The 11C seemed bulletproof, but this 32S is a bit flaky. Its the
    calculator I keep in my workshop drawer, so it's not used very often.
    But it seems that at least a third of the time I want to use it, it
    refuses to turn on. I've slipped a little note in its case with notes on >>> the recovery procedure.

    When it flakes out, I'll usually just pull out my Android phone and run
    the HP 48G emulator. My main complaint about that one is that it isn't
    really programmable - or at least, it doesn't retain programs when the
    app is killed.

    I finally remembered to look into my boxes of HP calculators. I don't
    have an HP 32SII or something comparable. In the scientific
    calculator section, I found an HP 31E and a 32E. Both of these are
    low end RPN calculators with LED displays. The 31E had a battery leak
    at some time in the past and will require that I do some battery
    contact rebuilding. I can provide a complete list of what I have in
    stock.

    I never could get used to the RPN data entry method

    Humans tend to prefer whatever technology they learned first. My
    first calculator was basically a mechanical "adding machine". <https://www.google.com/search?num=10&q=marchant+adding+machine&udm=2>
    I eventually ended up with an HP-35 RPN calculator and loved it. RPN
    is easy, if you think like a computer that stores intermediate results
    in a stack. At home, I use an HP-41CX or an emulator on my phone and
    PC. There are benefits and detriments to both algebraic and RPN
    notation. If we can become accustomed to QWERTY keyboards designed to
    slow down typing, we can get used to anything. I can switch back and
    forth between algebraic and RPN. Algebraic for financial calculations
    and RPN for engineering. Evaluating long equations is easier (for me)
    using RPN.


    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Shadow@21:1/5 to All on Fri Jan 10 13:28:26 2025
    On Fri, 10 Jan 2025 07:25:48 -0800, Jeff Liebermann <jeffl@cruzio.com>
    wrote:

    On Fri, 10 Jan 2025 05:19:53 -0500, zen cycle
    <funkmasterxx@hotmail.com> wrote:

    On 1/10/2025 12:20 AM, Jeff Liebermann wrote:
    On Sat, 4 Jan 2025 20:51:16 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/4/2025 7:28 PM, Jeff Liebermann wrote:

    I'll look in my collection and see if I have an HP11C. Offhand, I
    don't think so. There are some on eBay. The one's that offer the
    least risk and are likely to work are $75 and up.

    No need to check. When that calculator got stolen, the guys running the >>>> company bought me a replacement. By then, the 11C was no longer
    marketed, so I ended up with an HP 32S II, one of the few RPN machines >>>> still in their line, IIRC.

    The 11C seemed bulletproof, but this 32S is a bit flaky. Its the
    calculator I keep in my workshop drawer, so it's not used very often.
    But it seems that at least a third of the time I want to use it, it
    refuses to turn on. I've slipped a little note in its case with notes on >>>> the recovery procedure.

    When it flakes out, I'll usually just pull out my Android phone and run >>>> the HP 48G emulator. My main complaint about that one is that it isn't >>>> really programmable - or at least, it doesn't retain programs when the >>>> app is killed.

    I finally remembered to look into my boxes of HP calculators. I don't
    have an HP 32SII or something comparable. In the scientific
    calculator section, I found an HP 31E and a 32E. Both of these are
    low end RPN calculators with LED displays. The 31E had a battery leak
    at some time in the past and will require that I do some battery
    contact rebuilding. I can provide a complete list of what I have in
    stock.

    I never could get used to the RPN data entry method

    Humans tend to prefer whatever technology they learned first. My
    first calculator was basically a mechanical "adding machine". ><https://www.google.com/search?num=10&q=marchant+adding+machine&udm=2>
    I eventually ended up with an HP-35 RPN calculator and loved it. RPN
    is easy, if you think like a computer that stores intermediate results
    in a stack. At home, I use an HP-41CX or an emulator on my phone and
    PC. There are benefits and detriments to both algebraic and RPN
    notation. If we can become accustomed to QWERTY keyboards designed to
    slow down typing, we can get used to anything. I can switch back and
    forth between algebraic and RPN. Algebraic for financial calculations
    and RPN for engineering. Evaluating long equations is easier (for me)
    using RPN.

    There's a freeware HP-42S emulator here:

    <https://thomasokken.com/free42/>

    Just unzip the binary to a flashdrive (or whatever). If you
    put an empty file called "portable" in the same folder it won't write
    anything to your HD.

    I don't use it much, but it is very actively maintained. Last
    version was this year.
    []'s
    --
    Don't be evil - Google 2004
    We have a new policy - Google 2012
    Google Fuchsia - 2021

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Tom Kunich@21:1/5 to Roger Merriman on Fri Jan 10 18:58:41 2025
    On Fri, 03 Jan 2025 18:17:54 +0000, Roger Merriman wrote:

    Frank Krygowski <frkrygow@sbcglobal.net> wrote:
    On 1/3/2025 11:12 AM, cyclintom wrote:

    Coming out of Niles Canyon, you have to ride at around 20 mph Because
    of traffic I was forced to cross a rumble strip with my 28 mm tires
    and came damned close to losing control but it did allow me to let 5
    cars moving at 45 mph + get past before a constriction. While you're
    talking about taking the lane why don't you come here and try taking
    the lane? You would soon discover, if you're lucky, from a hospital
    bed that California deivers don't like your ideas.

    Ah. We haven't had a "Bicycling is really dangerous _HERE_!" post in
    quite a while.

    So you judged that nearly losing control in front of a 45 mph car was
    safer than legally taking the lane? Yes, my choice would have been
    different, and I've made that choice in <gasp!> California; but
    admittedly not in your super-dangerous neighborhood. When I do that,
    motorists wait until its safe to pass. Exceptions are vanishingly rare.

    I’d assume most folks would ie use the road than ride in the gutter on
    the rumble strips! Though I can’t see much evidence of any bar a central line, so as ever not sure how/why Tom would be riding in those.

    Seems on a very brief search that some gutters have been used by some cyclists as painted bike lanes, which isn’t a wildly good idea at best
    of times! And are unhappy at the possibility of encountering rumble
    strips, which seems likely to be a poor road all around!

    Do have some painted gutters though Heathrow which i suspect the might
    trick the unwary into thinking they are bike lanes, though it’s a fairly car centric type of roads so probably somewhat self selecting, ie I’ve
    only ever seen folks like myself ie brave folks on road bikes, though
    it’s a very rarely go though on the commute MTB which the gutters are
    less of no no as it’s plush tires are unfazed by drain covered, and one
    is traveling quite a lot slower, though even so it’s not a terribly wise idea.

    Roger Merriman

    As I often ask, what do you do when riding in a ten foot lane with no
    shoulder, when an 8.5 foot wide truck approaches from behind? Do you
    jump off the bike and humbly bow?







    Roger, I don't know where you're coming from. The rumble strip here isn't
    in the "gutter" but a 3" wide strip between the road and an open area you
    can pull over into - it allowed all of the traffic to pass without me even slowing down and the rumble strip disappeared allowing easy access to the
    open road before the constriction under a railroad overpass.

    This wouldn't be a subject save for Krygowski claiming that he ALWAYS took
    the lane.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Wolfgang Strobl@21:1/5 to All on Fri Jan 10 23:18:27 2025
    Am Fri, 10 Jan 2025 07:25:48 -0800 schrieb Jeff Liebermann
    <jeffl@cruzio.com>:

    On Fri, 10 Jan 2025 05:19:53 -0500, zen cycle
    <funkmasterxx@hotmail.com> wrote:

    On 1/10/2025 12:20 AM, Jeff Liebermann wrote:
    On Sat, 4 Jan 2025 20:51:16 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/4/2025 7:28 PM, Jeff Liebermann wrote:

    I'll look in my collection and see if I have an HP11C. Offhand, I
    don't think so. There are some on eBay. The one's that offer the
    least risk and are likely to work are $75 and up.

    No need to check. When that calculator got stolen, the guys running the >>>> company bought me a replacement. By then, the 11C was no longer
    marketed, so I ended up with an HP 32S II, one of the few RPN machines >>>> still in their line, IIRC.

    The 11C seemed bulletproof, but this 32S is a bit flaky. Its the
    calculator I keep in my workshop drawer, so it's not used very often.
    But it seems that at least a third of the time I want to use it, it
    refuses to turn on. I've slipped a little note in its case with notes on >>>> the recovery procedure.

    When it flakes out, I'll usually just pull out my Android phone and run >>>> the HP 48G emulator. My main complaint about that one is that it isn't >>>> really programmable - or at least, it doesn't retain programs when the >>>> app is killed.

    I finally remembered to look into my boxes of HP calculators. I don't
    have an HP 32SII or something comparable. In the scientific
    calculator section, I found an HP 31E and a 32E. Both of these are
    low end RPN calculators with LED displays. The 31E had a battery leak
    at some time in the past and will require that I do some battery
    contact rebuilding. I can provide a complete list of what I have in
    stock.

    I never could get used to the RPN data entry method

    Humans tend to prefer whatever technology they learned first.

    Not really. I learned programming by creating punched cards using an IBM
    26 Printing Card Punch, editing those programs by using the card
    duplicating feature of those machines. Followed by pinnig stretches of
    pages printed by large chain printers like to a wall, in order to study complaints from the Fortran IV compiler, or using a pen to mark logical
    errors in the source code, for later correction work on said card punch.

    <https://en.wikipedia.org/wiki/IBM_1403>

    I prefered almost every new programming technology that I learned later. Sometimes, technology development takes the wrong path, naturally we
    shouldn't and couldn't follow every idea, just because it is new. Simply because several competing ideas are usually in vogue and are cheered on
    by their enthusiastic fans.


    My
    first calculator was basically a mechanical "adding machine". ><https://www.google.com/search?num=10&q=marchant+adding+machine&udm=2>
    I eventually ended up with an HP-35 RPN calculator and loved it. RPN
    is easy, if you think like a computer that stores intermediate results
    in a stack.

    Of course. If you do. But while I did study the matter (computer
    science, that is), I don't think like a computer. I don't use RPN when
    writing mathematical formulas, neither on paper, nor while programming.
    Before you ask, I do know FORTH, I've even ported and changed an almost
    unknown FORTH like language, adding traditional operator precedences to
    the implementation, as a side project.

    There is nothing wrong with RPN, if it works for you. But there isn't
    anything "natural" about it.


    At home, I use an HP-41CX or an emulator on my phone and
    PC. There are benefits and detriments to both algebraic and RPN
    notation.

    Yep.

    I rarely used calculators, because writing short program snippets is
    more natural for me. Many decades ago I did that using PL/I-80 on my self-built CP/M computer, nowadays I just fire up a Python REPL, or
    start a Jupyter notebook, for so called "back of the envelope"
    calculations.

    I do have a calculator app on my phone, which doesn't emulate anything.
    It is switchable between traditional algebraic and RPN, radix modes
    etc., though. The most frequently used feature is the comprehensive
    conversion table for units, length, area, mass, speed and such. How
    would I be able to convert Fathoms to Furlongs, without such a tool? :-)


    If we can become accustomed to QWERTY keyboards designed to
    slow down typing, we can get used to anything. I can switch back and
    forth between algebraic and RPN. Algebraic for financial calculations
    and RPN for engineering. Evaluating long equations is easier (for me)
    using RPN.

    I can relate to that, but personally, I just don't calculate that way, I
    don't do one shot calculations often enough to earn a benefit from RPN.

    My first and only HP calculator is a HP 200LX, still working fine. I
    lost one of the tiny case screws, decades ago, replacing the CR2032 now
    and then is still awkward. I have no real use for the device anymore,
    but I still like it enough for not giving it away. AFAIR, there's still
    a copy of the original DeSmet C compiler on the flash card in the PCMCIA
    slot. :-)


    Getting back to cycling ... :-)

    27 years ago, in 1998, the HP 200LX spent a lot of time in my panniers
    on my commute (~5000 km/a). While playing with and exploring tiny
    Microchip PIC microcontrollers at that time as a hobby, I used a spare
    and slightly defective homebrew PCB (size ~1,3 *1,4 inch) as a prototype
    board, took a PC16C84 processor, a 4,194,304 Hz quartz, 3 small
    capacitors, one resistor, a 9 pin RS232 connector and two sockets, in
    order to build an interface between the HPLX200 and a standard reed
    contact as used by most bicycle computers sold at that time.

    For simplicity, I didn't implement anything fancy on the PIC, by just
    measuring and counting revolutions for an adjustable time span and then
    sending the data in plain ASCII via the standard null modem cable which
    came with the 200LX, leaving collecting the data and perhaps doing the calculations on the 200LX.

    I worked quite well, for some definition of "well". As often, software
    and electronics were not the problem. The mechanics were the problem,
    more precisely, the proprietary cable from the HP to my circuit board.
    So in the end, I just collected data from a single test drive by using
    the log feature of the 200LXs terminal program, while carefully holding
    the plug with one hand, transfered that to a PC, used Excel to produce a
    table - and finally admired my work. :-)

    *** from PIC *******
    One Revolution s
    Current speed km/s
    Accum. rev #
    Distance km
    Time s
    800 6 2531 0,62 12,5 6 0,013 50
    880 9 1867 0,46 17 15 0,032 55
    960 16 1174 0,29 27 31 0,067 60
    1040 19 1123 0,27 28,2 50 0,108 65
    1120 17 1186 0,29 26,7 67 0,144 70
    1200 16 1238 0,3 25,6 83 0,178 75
    1280 15 1311 0,32 24,2 98 0,211 80
    1360 14 1400 0,34 22,6 112 0,241 85
    1440 14 1374 0,34 23,1 126 0,271 90
    1520 17 1209 0,3 26,2 143 0,307 95
    1600 16 1309 0,32 24,2 159 0,342 100
    1680 15 1426 0,35 22,2 174 0,374 105
    1760 14 1458 0,36 21,7 188 0,404 110
    1840 14 1474 0,36 21,5 202 0,434 115
    1920 14 1456 0,36 21,8 216 0,464 120
    2000 14 1455 0,36 21,8 230 0,495 125
    2080 15 1432 0,35 22,1 245 0,527 130
    2160 14 1450 0,35 21,9 259 0,557 135
    2240 14 1461 0,36 21,7 273 0,587 140
    2320 14 1471 0,36 21,6 287 0,617 145
    2400 14 1560 0,38 20,3 301 0,647 150


    Anyway, having worked exclusively with mainframes in my job during the
    early years, followed by PC sized smaller computers connected to what
    they call "cloud" these days, I developed an interest in tiny
    microcontrollers that were actually simple and useable for a hobbyist.
    This started with the early PIC from Microchip, still expensive, needing expensive UV-erasable versions, followed by the PIC 18C84 EEPROM
    version, which could be reprogrammed without needing an somewhat
    expensive UV lamp. Even more joy was using much smaller 8 bit
    controllers that were still able to do work that what would have needed
    a lot of TTL logic, beforehand. Like for example the PIC10F200 <https://ww1.microchip.com/downloads/aemDocuments/documents/OTH/ProductDocuments/DataSheets/40001239F.pdf>
    Harvard architecture, DIP8, 4 MHz, 256 words program memory, 16 Bytes
    RAM, 3 I/O pins, one Input only pin, internal pullups, serial
    programmable using an inexpensive programmer. One doesn't need an ARM
    CPU with a handfull of cores in order to operate a mid sized toy
    railroad, a handfull of these tiny and cheap microcontrollers would do,
    at least in principle.

    There are cheaper microcontrollers of Chinese origin, Padauk comes to
    mind. A quick search tells the there are still cheaper ones in the
    pipeline (1.5 cents for a complete computer, wow!) <https://zeptobars.com/en/read/Nyquest-Technology-NY8A051H-8051-smallest-microcontroller>
    but frankly, that's something for the next generation to play with. Is
    it possible to implement Tetris with 1k words of OTP memory and 48 bytes
    of SRAM? What about a Mastermind solver? I implemented a
    Mastermind/Bulls&Cows solver using the PIC16C84, complete with decoding
    two buttons, driving/multiplexing a raw four-digit seven-segment
    display, using most of its 2 K words of EEPROM and 68 bytes of RAM.


    <https://www.mystrobl.de/ws/pic/mm47/>

    Parts are still available in 2025, including PIC 16F84A (€6.30 at a
    local dealer) and HDSP 5503 (between 4 and 8 $ at ebay), so building
    that gadget would still be possible. Rather uninteresting, though,
    because microelectronics came a long way between the 16C84 and what we
    can use buy and even as hobbyists, today.


    This makes me reflect on the criticism of electronics in bicycles. I'm
    not talking about replacing muscle power by motor power, that's
    replacing bicycles by something else. But what about measuring the
    amount of power applied to the pedals, what about telling the cyclist
    who balanced he splits the power between left and right, by
    instrumenting the pedal or the bottom bracket? What about replacing
    those awkward cables and complicated brifters with simple electric
    switches and an encrypted wireless channel? Is that bad, because a
    blacksmith can't repair it with his tools, like giving a horse a new
    pair of horse shoes? What about LED lights, then? Shouldn't we get
    back to incandescent bulbs, powered by bottle dynamos?

    I think we should not and can not turn back the wheel, at least not like
    this. I accept that there are reasons to keep bicycles simple, or to
    keep at least some biycles simple enough to long lived and usable even
    without much maintenance and without exotic stuff. But the question is,
    what makes a component or material exotic? Is a specific bowden cable or
    a gear hub or hub generator really less exotic and simpler to
    replace/recreate than, say, a LED light or a wireless shifter? I doubt
    it.

    There is a point when mechanical parts become complicated enough to
    create a vendor lock-in, when a second source isn't available. Just like
    with electronic parts. Sometimes the relationship gets reversed, when an over-engineered and complicated mechanical solution is replaced by a
    simple construction that combines commercially available electronics
    with simple mechanical parts.

    In this respect, you have to take a close look at where a dependency
    arises, instead of simply linking it to characteristics such as “new”, “electronic” or “wireless”. We need open standards, either through industry commitments, or by regulations.



    --
    Thank you for observing all safety precautions

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From zen cycle@21:1/5 to Wolfgang Strobl on Fri Jan 10 23:57:20 2025
    On 1/10/2025 5:18 PM, Wolfgang Strobl wrote:
    <BIG snip>


    This makes me reflect on the criticism of electronics in bicycles. I'm
    not talking about replacing muscle power by motor power, that's
    replacing bicycles by something else. But what about measuring the
    amount of power applied to the pedals, what about telling the cyclist
    who balanced he splits the power between left and right, by
    instrumenting the pedal or the bottom bracket? What about replacing
    those awkward cables and complicated brifters with simple electric
    switches and an encrypted wireless channel? Is that bad, because a
    blacksmith can't repair it with his tools, like giving a horse a new
    pair of horse shoes? What about LED lights, then? Shouldn't we get
    back to incandescent bulbs, powered by bottle dynamos?

    I think we should not and can not turn back the wheel, at least not like this. I accept that there are reasons to keep bicycles simple, or to
    keep at least some biycles simple enough to long lived and usable even without much maintenance and without exotic stuff. But the question is,
    what makes a component or material exotic? Is a specific bowden cable or
    a gear hub or hub generator really less exotic and simpler to replace/recreate than, say, a LED light or a wireless shifter? I doubt
    it.

    There is a point when mechanical parts become complicated enough to
    create a vendor lock-in, when a second source isn't available. Just like
    with electronic parts. Sometimes the relationship gets reversed, when an over-engineered and complicated mechanical solution is replaced by a
    simple construction that combines commercially available electronics
    with simple mechanical parts.

    In this respect, you have to take a close look at where a dependency
    arises, instead of simply linking it to characteristics such as “new”, “electronic” or “wireless”. We need open standards, either through industry commitments, or by regulations.

    +1
    A well-written pro technology rant.
    thank you.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From zen cycle@21:1/5 to Wolfgang Strobl on Fri Jan 10 23:47:44 2025
    On 1/10/2025 5:18 PM, Wolfgang Strobl wrote:
    Am Fri, 10 Jan 2025 07:25:48 -0800 schrieb Jeff Liebermann <jeffl@cruzio.com>:

    On Fri, 10 Jan 2025 05:19:53 -0500, zen cycle
    <funkmasterxx@hotmail.com> wrote:

    On 1/10/2025 12:20 AM, Jeff Liebermann wrote:
    On Sat, 4 Jan 2025 20:51:16 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/4/2025 7:28 PM, Jeff Liebermann wrote:

    I'll look in my collection and see if I have an HP11C. Offhand, I >>>>>> don't think so. There are some on eBay. The one's that offer the >>>>>> least risk and are likely to work are $75 and up.

    No need to check. When that calculator got stolen, the guys running the >>>>> company bought me a replacement. By then, the 11C was no longer
    marketed, so I ended up with an HP 32S II, one of the few RPN machines >>>>> still in their line, IIRC.

    The 11C seemed bulletproof, but this 32S is a bit flaky. Its the
    calculator I keep in my workshop drawer, so it's not used very often. >>>>> But it seems that at least a third of the time I want to use it, it
    refuses to turn on. I've slipped a little note in its case with notes on >>>>> the recovery procedure.

    When it flakes out, I'll usually just pull out my Android phone and run >>>>> the HP 48G emulator. My main complaint about that one is that it isn't >>>>> really programmable - or at least, it doesn't retain programs when the >>>>> app is killed.

    I finally remembered to look into my boxes of HP calculators. I don't >>>> have an HP 32SII or something comparable. In the scientific
    calculator section, I found an HP 31E and a 32E. Both of these are
    low end RPN calculators with LED displays. The 31E had a battery leak >>>> at some time in the past and will require that I do some battery
    contact rebuilding. I can provide a complete list of what I have in
    stock.

    I never could get used to the RPN data entry method

    Humans tend to prefer whatever technology they learned first.

    Not really. I learned programming by creating punched cards using an IBM
    26 Printing Card Punch, editing those programs by using the card
    duplicating feature of those machines. Followed by pinnig stretches of
    pages printed by large chain printers like to a wall, in order to study complaints from the Fortran IV compiler, or using a pen to mark logical errors in the source code, for later correction work on said card punch.

    MY freshman year at Northeastern was the first year they didn't use
    punchcards for the intro fortran class. We were 'treated' to the TA
    giving a punchcard demonstration at one point. I uttered a prayer of
    thanks to the Technology Gods.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From zen cycle@21:1/5 to Frank Krygowski on Sat Jan 11 00:06:46 2025
    On 1/10/2025 8:01 PM, Frank Krygowski wrote:
    On 1/10/2025 5:18 PM, Wolfgang Strobl wrote:

    This makes me reflect on the criticism of electronics in bicycles. ...
    But what about measuring the
    amount of power applied to the pedals, what about telling the cyclist
    who balanced he splits the power between left and right, by
    instrumenting the pedal or the bottom bracket?

    Yes, that can be done with enough sensors and electronics. But it seems
    like useless information to me. As with much computer software, it seems
    like "feature bloat." Why would anybody but a racer care?

    One doesn't need to be a racer to be interested in those types of
    metrics. I know plenty of "former" racers, and "sport" cyclists that
    keep diligent track of their workouts.


    What about replacing
    those awkward cables and complicated brifters with simple electric
    switches and an encrypted wireless channel? Is that bad, because a
    blacksmith can't repair it with his tools, like giving a horse a new
    pair of horse shoes?

    Of course, you're welcome to use electric shifting if you like. (Or
    brifters, which I don't!) But my life experience with electronic devices makes me suspicious. I've seen too many examples of electronic devices
    that simply stopped working, with no possible way of diagnosing the
    problem - at least, not by me. And while I'm far from an electronics
    expert, I'm better than the average citizen.

    I'm no fan of electronic shifting - it would be just my luck to be on
    planed 4-6 hour ride, and have something crap out at the furthest point
    from the car/house. That said, I'm never going back to discrete shifters.


    Two days ago, my kid asked me to figure out why an electric blanket
    wasn't working. The controller refused to turn on. I opened it and
    confirmed that it was getting supply voltage. Beyond that, the pile of
    dozens of surface mount electronic components was incomprehensible to
    me. I suspect Jeff might have been able to diagnose it, but not me.
    Here's a photo: https://www.flickr.com/photos/16972296@N08/54259119364/in/dateposted-
    public/
    I know electric blankets are old technology. I know they functioned well
    for decades with maybe a rheostat and perhaps a couple of other
    components. Why add unrepairable complexity?

    Becasue it's quite a bit cheaper to produce, and generally works just as
    well.


    Similarly, a good friend recently told me about her adventure with her relatively new washing machine. During an expensive service visit, a repairman told her the problem was somewhere in the main circuit board,
    and that the only solution was to replace the expensive board. Our
    washer is something like 35 years old, uses an electro-mechanical timer,
    and will probably work well for whichever grandkid inherits it.

    I'm positive my shifters are also going to be working for whichever
    grandkid inherits them.

    I'm just as optimistic about my Sram integrated system, but the idea
    that they would see anything a=of value in my collection of bikes...I'm
    kinda skeptical about.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Wolfgang Strobl@21:1/5 to All on Sat Jan 11 22:11:57 2025
    Am Fri, 10 Jan 2025 20:01:26 -0500 schrieb Frank Krygowski <frkrygow@sbcglobal.net>:

    On 1/10/2025 5:18 PM, Wolfgang Strobl wrote:

    This makes me reflect on the criticism of electronics in bicycles. ... But what about measuring the
    amount of power applied to the pedals, what about telling the cyclist
    who balanced he splits the power between left and right, by
    instrumenting the pedal or the bottom bracket?

    Yes, that can be done with enough sensors and electronics. But it seems
    like useless information to me. As with much computer software, it seems
    like "feature bloat." Why would anybody but a racer care?

    Is there any reason to believe that old people like me don't have a need
    to care about training intensity, as a matter of principle? How comes?

    When using my biycles as a middle aged, healthy adult I didn't care
    about racing, didn't train and didn't try to find company in cycling
    clubs, either. I didn't visit a gym until much later. Cycling was simply
    a way to get to work quickly. During vacations it was a more
    entertaining mode of transportation than the car. Fitness, strength and
    cardio fitness came as a side effect, slowly and almost unnoticed.

    This has changed. My cardio fitness is still better than what I know
    about a lot of people half my age. Nevertheless, it decreased over the
    years and it takes more and more effort to keep what you have not yet
    lost. As it is expected. The ability to measure power is helpful in
    finding the point at which intensity is sufficient and overload has not
    yet begun. This is even more true in combination with a heart rate
    monitor.

    So far, this is only about aging. But there is more.

    Accidents that led to damage to bones, joints and tendons in the past
    have consequences to be considered, too. There is a tendency to
    compensate weaknesses by bad postures, whithout noticing. This is both
    caused by damages and causing damages. This is to be avoided. Measuring
    how power is applied by the feet helps detecting differences early,
    avoiding damage.

    Avoiding damage is far from useless.



    What about replacing
    those awkward cables and complicated brifters with simple electric
    switches and an encrypted wireless channel? Is that bad, because a
    blacksmith can't repair it with his tools, like giving a horse a new
    pair of horse shoes?

    Of course, you're welcome to use electric shifting if you like.

    Thank you very much. :-)


    (Or
    brifters, which I don't!) But my life experience with electronic devices >makes me suspicious. I've seen too many examples of electronic devices
    that simply stopped working, with no possible way of diagnosing the
    problem - at least, not by me. And while I'm far from an electronics
    expert, I'm better than the average citizen.

    Live and learn.

    My experiences with mechanical devices are at least as mixed as those
    with electronic devices. Some very complicated electronic devices that
    I bought decades ago still work. The same applies to some that I built
    myself a long time ago. With devices that contained both electronic and mechanical parts, the fault was more often on the mechanical side.

    Anecdotical, I know.


    Two days ago, my kid asked me to figure out why an electric blanket
    wasn't working. The controller refused to turn on. I opened it and
    confirmed that it was getting supply voltage. Beyond that, the pile of
    dozens of surface mount electronic components was incomprehensible to
    me. I suspect Jeff might have been able to diagnose it, but not me.
    Here's a photo: >https://www.flickr.com/photos/16972296@N08/54259119364/in/dateposted-public/ >I know electric blankets are old technology. I know they functioned well
    for decades with maybe a rheostat and perhaps a couple of other
    components. Why add unrepairable complexity?

    Don't know. We use devices like the blue one in the following picture. <https://upload.wikimedia.org/wikipedia/commons/4/4b/W%C3%A4rmflasche1.jpg>
    and better isolating blankets. No electricity necessary. :-)

    A few lights in our house are switched by set of 2 x 3 inexpensive
    wireless sockets including two remote controls, that I bought eleven
    years ago. I've still to replace the batteries. Two of the sockets are
    still spares, I have a replacement cell for the remote controls stored
    which might live even longer. Standard type, used in garage openers and
    burglar alarms, too. Selecting a channel and paring one of the four
    buttons of a remote control with one or more of the sockets is as easy
    as pie, using a line of dip switches inside those devices. Quite
    similar to pairing switches and derailleur on our bicyles.

    While I avoid having essential functionality in my house depend on
    wireless connections, I enjoy having the option, for certain use cases,
    though. Same with bicycles.



    Similarly, a good friend recently told me about her adventure with her >relatively new washing machine. During an expensive service visit, a >repairman told her the problem was somewhere in the main circuit board,
    and that the only solution was to replace the expensive board. Our
    washer is something like 35 years old, uses an electro-mechanical timer,
    and will probably work well for whichever grandkid inherits it.

    I can relate to that, up to a point. A lot of our equipment is or was
    of similar age, or even older. Had a similar experience with our oven,
    when it was ten years and a month old. Error code xyz, "call service". Expensive service visit by NEFF, proposed solution was replacing thre
    expensive boards one after the other, without any guarantee. Almost as expensive as a new oven. Refused, payed the service, decided not to buy anything from NEFF again and tried the repair myself. Found a pair of
    cold solder joints just before the heating coil causing some protection
    circuit to trigger, resoldered. Problem solved. Wondered why a cold
    joint took so long to break, then remembered having replaced the heating
    coil years ago because of intermittent problems. Most probably, I
    replaced a working heating coil, accidently somehow fixing the
    connection, perhaps just by moving the oven out of his compartement. I'm
    still using that oven for baking bread every two or three day.

    On the other hand, while I won't enjoy having to replace our old washing machine and the dryer, I know that new machines are a lot more efficent.
    Dryers produce water instead of releasing hot steam, washing machines
    spin the loundry much faster, etc.



    I'm positive my shifters are also going to be working for whichever
    grandkid inherits them. I wouldn't be confident about electronic shifters.

    I still have some carpentry tools which where already old before I was
    born. I don't throw them away out of respect, but don't use any of
    these.

    In a world where people get rid of cars when the are a few years old, I
    don't see a point in conserving antique bicyles for later generations.
    Not for _using_ these bicycles, that is. I'm quite confident that it is
    easier to make a electronic derailleur outlive the bicycle than doing
    that with a mechanical one. Don't know about actual product, though. I
    have my doubts about the shifter, though. Guess what: it is a mechanical problem. Replacing the CR2032 without damaging the contacts is a
    mechanical problem. Cold be solved by a better mechanical solution, but
    not by me.


    What about LED lights, then? Shouldn't we get
    back to incandescent bulbs, powered by bottle dynamos?

    I love good LED bike lights. And in some cases, the LEDs are direct >replacements for incandescent bulbs.

    Eeek. I'd rather have a LED built into a well designed free form
    reflector.

    But I still use bottle dynos on two
    bikes. They can be more than fine, depending on service requirements.

    Sure. Exceptions from the rule are to be expected.


    I think we should not and can not turn back the wheel, at least not like
    this. I accept that there are reasons to keep bicycles simple, or to
    keep at least some biycles simple enough to long lived and usable even
    without much maintenance and without exotic stuff. But the question is,
    what makes a component or material exotic? Is a specific bowden cable or
    a gear hub or hub generator really less exotic and simpler to
    replace/recreate than, say, a LED light or a wireless shifter? I doubt
    it.

    It depends on the LED light, I suppose. Some seem to have on board >electronics as complicated as that blanket controller,

    So what. At a certain point electronics just are components. You
    wouldn't repair a halogen bulb, either. I don't care about how
    complicated an electronic board is, I care about its MTBF.


    and I don't know
    why. (I am curious about that, having opened up one Busch & Muller
    headlamp.) But I've opened and repaired both gear hubs and hub dynos
    from the 1950s. They now work perfectly. That won't be true of current >electronic shifters 50 years from now.

    I broke two gear hubs in sequence, didn't repair either one myself,
    because I didn't have time for that, because of job, family and the time
    needed for a long commute by bike. First one broken beyond repair,
    according to the shop, I learned to live with a missing gear with the
    second one. Bicycles are tools which wear out both through use and technological progress. A lot sold as progress is just fashion. But
    sometimes, a change is progress. I like having choices.

    When I see a 50 year old, perfect looking bicycle, I think:
    That one must have been standing around inside most of the time, perhaps
    for a reason. There actually is such a bicycle in our family, a local
    shop built it for my wife, long ago. Served its purpose, it is not
    nearly as old, but old and nice enough to find a place to hang it on the
    wall. But taking it for a ride? Perhaps, for riding to a nearby ice
    cream parlor during summer. :-)

    --
    Thank you for observing all safety precautions

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Wolfgang Strobl@21:1/5 to All on Sat Jan 11 22:22:07 2025
    Am Sat, 11 Jan 2025 00:06:46 -0500 schrieb zen cycle <funkmasterxx@hotmail.com>:


    I'm no fan of electronic shifting - it would be just my luck to be on
    planed 4-6 hour ride, and have something crap out at the furthest point
    from the car/house. That said, I'm never going back to discrete shifters.

    It isn't that difficult to carry a CR2032 and a tiny 25 g rechargeable
    battery somewhere. The coin cell lasts about two years, the battery is
    good for about 800-1000 km. Both warn early via LED on the device or via
    a Garmin Edge computer, for not having to care during a 4-6 hour ride.

    --
    Thank you for observing all safety precautions

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Catrike Ryder@21:1/5 to frkrygow@sbcglobal.net on Sun Jan 12 04:33:25 2025
    On Sat, 11 Jan 2025 19:47:52 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/11/2025 4:22 PM, Wolfgang Strobl wrote:
    Am Sat, 11 Jan 2025 00:06:46 -0500 schrieb zen cycle
    <funkmasterxx@hotmail.com>:


    I'm no fan of electronic shifting - it would be just my luck to be on
    planed 4-6 hour ride, and have something crap out at the furthest point >>>from the car/house. That said, I'm never going back to discrete shifters.

    It isn't that difficult to carry a CR2032 and a tiny 25 g rechargeable
    battery somewhere. The coin cell lasts about two years, the battery is
    good for about 800-1000 km. Both warn early via LED on the device or via
    a Garmin Edge computer, for not having to care during a 4-6 hour ride.


    But how many threads have we had here about problems with Garmin computers?

    Not so many. I did have the Garmin Edge 130+ computer fail, but I
    don't recall any other Garmin computer failures reported.

    I also have a Garmin Edge 500 that I bought in 2011 and retired in
    2023 that still works perfectly. I upgraded to get some additional
    features. My wife uses it currently on her Catrike Pocket which is
    inside on rollers.

    --
    C'est bon
    Soloman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From zen cycle@21:1/5 to Wolfgang Strobl on Sun Jan 12 07:22:01 2025
    On 1/11/2025 4:22 PM, Wolfgang Strobl wrote:
    Am Sat, 11 Jan 2025 00:06:46 -0500 schrieb zen cycle <funkmasterxx@hotmail.com>:


    I'm no fan of electronic shifting - it would be just my luck to be on
    planed 4-6 hour ride, and have something crap out at the furthest point >>from the car/house. That said, I'm never going back to discrete shifters.

    It isn't that difficult to carry a CR2032 and a tiny 25 g rechargeable battery somewhere. The coin cell lasts about two years, the battery is
    good for about 800-1000 km. Both warn early via LED on the device or via
    a Garmin Edge computer, for not having to care during a 4-6 hour ride.


    I can't tell you how many times I leave the house and forget my wallet.
    Adding trying to remember to carry a spare battery is only part of the
    problem though.

    The other issue is the failure rates associated with a complex electro-mechanical systems VS a simple cable derailleur system. (Since I conduct FMEDAs as part of my role, I'm well-aware of the exponential
    failure rates associated with increased complexity).

    Since I've been seriously riding in the early 1980's, I've never had a
    problem associated with a shifting mechanism that I couldn't handle with
    a simple road/trail side repair, except for the few times when I've
    sheared the derailleur off in the woods. If your E-shifting mechanism
    fails for any reason other than the battery going dead, you're pretty
    much stick in whenever gear it left you in.

    I don't begrudge anyone for wanting E-shifting. I think it's a great technological advancement. The few times I've ridden E-shifting systems,
    I've been impressed with the accuracy, repeatability, and quickness of
    the shifts. But I don't see those advantages of being worth the extra
    cost and risk of failure. That's just my opinion, YMMV.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From zen cycle@21:1/5 to Frank Krygowski on Sun Jan 12 07:49:31 2025
    On 1/11/2025 7:46 PM, Frank Krygowski wrote:
    On 1/11/2025 4:11 PM, Wolfgang Strobl wrote:
    Am Fri, 10 Jan 2025 20:01:26 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    On 1/10/2025 5:18 PM, Wolfgang Strobl wrote:
    This makes me reflect on the criticism of electronics in
    bicycles. ... But what about measuring the
    amount of power applied to the pedals, what about telling the cyclist
    who balanced he splits the power between left and right, by
    instrumenting the pedal or the bottom bracket?

    Yes, that can be done with enough sensors and electronics. But it seems
    like useless information to me. As with much computer software, it seems >>> like "feature bloat." Why would anybody but a racer care?

    Is there any reason to believe that old people like me don't have a need
    to care about training intensity, as a matter of principle?  How comes?

    When using my biycles as a middle aged, healthy adult I didn't care
    about racing, didn't train and didn't try to find company in cycling
    clubs, either. I didn't visit a gym until much later. Cycling was simply
    a way to get to work quickly. During vacations it was a more
    entertaining mode of transportation than the car. Fitness, strength and
    cardio fitness came as a side effect, slowly and almost unnoticed.

    This has changed. My cardio fitness is still better than what I know
    about a lot of people half my age. Nevertheless, it decreased over the
    years and it takes more and more effort to keep what you have not yet
    lost. As it is expected. The ability to measure power is helpful in
    finding the point at which intensity is sufficient and overload has not
    yet begun. This is even more true in combination with a heart rate
    monitor.

    So far, this is only about aging. But there is more.

    Accidents that led to damage to bones, joints and tendons in the past
    have consequences to be considered, too. There is a tendency to
    compensate weaknesses by bad postures, whithout noticing. This is both
    caused by damages and causing damages. This is to be avoided.  Measuring
    how power is applied by the feet helps detecting differences early,
    avoiding damage.

    Avoiding damage is far from useless.

    But individual power outputs from left and right legs? I can't imagine
    much value from that measurement.

    You answered your own question. "You" can't imagine it's useful. There
    are a great many data geeks who find it quite useful.

    In fact, I don't believe power
    measurement is necessary for fitness. I judge by feel. It has the
    advantage of naturally compensating for days when I feel stronger or
    weaker. And it's organic!  ;-)

    In the exercise physiology world that's called Rate of Perceived
    Exertion (RPE) and is an important factor, but it's vague, difficult to
    assign any metric to, and only really detectable on a macro scale. If
    you really want to improve fitness beyond just feeling better, you need repeatable metrics with reasonable accuracy and resolution. RPE ain't it.



    My experiences with mechanical devices are at least as mixed as those
    with electronic devices.  Some very complicated electronic devices that
    I bought decades ago still work. The same applies to some that I built
    myself a long time ago.  With devices that contained both electronic and
    mechanical parts, the fault was more often on the mechanical side.

    Anecdotical, I know.

    To me, a big advantage is the ability to _look_ at a mechanical device
    and _see_ what's wrong. That, and the fact I can often affect a repair.
    I hate the Kleenex ethic - "It's no good any more, just throw it away."

    I agree with this to the extent that from a personal perspective I don't
    see the added value. However, I fully support the development for those
    that do find it useful.


    Two days ago, my kid asked me to figure out why an electric blanket
    wasn't working. The controller refused to turn on. I opened it and
    confirmed that it was getting supply voltage. Beyond that, the pile of
    dozens of surface mount electronic components was incomprehensible to
    me. I suspect Jeff might have been able to diagnose it, but not me.
    Here's a photo:
    https://www.flickr.com/photos/16972296@N08/54259119364/in/dateposted-
    public/
    I know electric blankets are old technology. I know they functioned well >>> for decades with maybe a rheostat and perhaps a couple of other
    components. Why add unrepairable complexity?

    Don't know.  We use devices like the blue one in the following picture.
    <https://upload.wikimedia.org/wikipedia/commons/4/4b/
    W%C3%A4rmflasche1.jpg>
    and better isolating blankets. No electricity necessary. :-)

    :-) But you imply that _I'm_ the one insufficiently modern?

    A few lights in our house are switched by set of 2 x 3 inexpensive
    wireless sockets including two remote controls, that I bought eleven
    years ago. I've still to replace the batteries. Two of the sockets are
    still spares, I have a replacement cell for the remote controls stored
    which might live even longer. Standard type, used in garage openers and
    burglar alarms, too. Selecting a channel and paring one of the four
    buttons of a remote control with one or more of the sockets is as easy
    as pie, using a line of dip switches inside those devices.  Quite
    similar to pairing switches and derailleur on our bicyles.

    While I avoid having essential functionality in my house depend on
    wireless connections, I enjoy having the option, for certain use cases,
    though.

    Yuck.

    I bought a new pedestal fan for our back patio last year. I've found a
    fan to be the best defense against mosquitos. The previous one still
    runs, but exposure to sunlight destroyed its plastic finger protection screen.

    Anyway, the new one comes with a remote control, roughly the size of a
    snack cracker. Who needs a remote to adjust a fan? I rejected ceiling
    fans with remotes when I was shopping, too. It's something to lose,
    something whose battery will die, and something whose electronics will
    go bad and be unrepairable.
    ...

    When I see a 50 year old, perfect looking bicycle, I think:
    That one must have been standing around inside most of the time, perhaps
    for a reason.

    My favorite bike is a young 39 years. It looks perfect, in part because
    I had it (and my wife's matching one) stripped and powder coated a
    couple years ago. Too many excellent memories to let it go.

    Or, if you're like me, I had my first racing bike fully restored, and
    now it hangs in my hallway on display hooks with the fork crown at
    eye-level. Some people really enjoy "bike porn": https://veloclassics.blogspot.com/




    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Wolfgang Strobl@21:1/5 to All on Sun Jan 12 15:54:52 2025
    Am Sat, 11 Jan 2025 19:47:52 -0500 schrieb Frank Krygowski <frkrygow@sbcglobal.net>:

    On 1/11/2025 4:22 PM, Wolfgang Strobl wrote:
    Am Sat, 11 Jan 2025 00:06:46 -0500 schrieb zen cycle
    <funkmasterxx@hotmail.com>:


    I'm no fan of electronic shifting - it would be just my luck to be on
    planed 4-6 hour ride, and have something crap out at the furthest point >>>from the car/house. That said, I'm never going back to discrete shifters.

    It isn't that difficult to carry a CR2032 and a tiny 25 g rechargeable
    battery somewhere. The coin cell lasts about two years, the battery is
    good for about 800-1000 km. Both warn early via LED on the device or via
    a Garmin Edge computer, for not having to care during a 4-6 hour ride.


    But how many threads have we had here about problems with Garmin computers?

    Don't know. My Garmin GPSMap 60CSx <https://www.mystrobl.de/Plone/radfahren/technik/komponenten/navi/IMG_1405.jpeg>
    that I bought in early 2008 is still working, but I replaced it by a
    better version (a 64 s), when this was sold at a very low price, in
    2018. I still use it as an outdoor device, when walking or for just
    recording a track on a long car ride. Can't report real problems when
    using an EDGE 1030 plus, so far, because I haven't had none.

    By the way, before it gets lost, while it is indeed convenient to be
    warned about low battery by the Garmin 1030, it isn't necessary for
    checking battery state. Both shifters display the state, when activated
    (full, low, critical), and so does the derailleur. In both cases, low is signaled early enough to still have about a day of riding left.

    There _are_ problems with devices like that, just with any new
    technology. But these aren't. Our bikes and their wireless shifting is perfectly useable without a Garmin computer, or a similar device of
    another brand.


    In my home town, I see many cyclists riding around on almost flat rear
    tires. This is somewhat dangerous. New cars have obligatory tire
    pressure warning, for quite some time. Most of these need maintenance,
    costing money. Such devices exist for bicycles, too. One of our sons
    used such a device on his bike, out of curiosity. It caused more
    problems than it solved, so he dismantled it.

    I fear that some fool comes up with the idea of making such a
    montitoring device mandatory for bicycles, too, perhaps by pointing to
    modern bicycles, ahem, to E-Bikes. :-/ We can't be safe enough, right?
    Right? :-)


    --
    Thank you for observing all safety precautions

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Frank Krygowski on Sun Jan 12 11:08:45 2025
    On 1/11/2025 6:46 PM, Frank Krygowski wrote:
    On 1/11/2025 4:11 PM, Wolfgang Strobl wrote:
    Am Fri, 10 Jan 2025 20:01:26 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    On 1/10/2025 5:18 PM, Wolfgang Strobl wrote:
    This makes me reflect on the criticism of electronics in
    bicycles. ... But what about measuring the
    amount of power applied to the pedals, what about
    telling the cyclist
    who balanced he splits the power between left and right, by
    instrumenting the pedal or the bottom bracket?

    Yes, that can be done with enough sensors and
    electronics. But it seems
    like useless information to me. As with much computer
    software, it seems
    like "feature bloat." Why would anybody but a racer care?

    Is there any reason to believe that old people like me
    don't have a need
    to care about training intensity, as a matter of
    principle?  How comes?

    When using my biycles as a middle aged, healthy adult I
    didn't care
    about racing, didn't train and didn't try to find company
    in cycling
    clubs, either. I didn't visit a gym until much later.
    Cycling was simply
    a way to get to work quickly. During vacations it was a more
    entertaining mode of transportation than the car. Fitness,
    strength and
    cardio fitness came as a side effect, slowly and almost
    unnoticed.

    This has changed. My cardio fitness is still better than
    what I know
    about a lot of people half my age. Nevertheless, it
    decreased over the
    years and it takes more and more effort to keep what you
    have not yet
    lost. As it is expected. The ability to measure power is
    helpful in
    finding the point at which intensity is sufficient and
    overload has not
    yet begun. This is even more true in combination with a
    heart rate
    monitor.

    So far, this is only about aging. But there is more.

    Accidents that led to damage to bones, joints and tendons
    in the past
    have consequences to be considered, too. There is a
    tendency to
    compensate weaknesses by bad postures, whithout noticing.
    This is both
    caused by damages and causing damages. This is to be
    avoided.  Measuring
    how power is applied by the feet helps detecting
    differences early,
    avoiding damage.

    Avoiding damage is far from useless.

    But individual power outputs from left and right legs? I
    can't imagine much value from that measurement. In fact, I
    don't believe power measurement is necessary for fitness. I
    judge by feel. It has the advantage of naturally
    compensating for days when I feel stronger or weaker. And
    it's organic!  ;-)

    My experiences with mechanical devices are at least as
    mixed as those
    with electronic devices.  Some very complicated electronic
    devices that
    I bought decades ago still work. The same applies to some
    that I built
    myself a long time ago.  With devices that contained both
    electronic and
    mechanical parts, the fault was more often on the
    mechanical side.

    Anecdotical, I know.

    To me, a big advantage is the ability to _look_ at a
    mechanical device and _see_ what's wrong. That, and the fact
    I can often affect a repair. I hate the Kleenex ethic -
    "It's no good any more, just throw it away."

    Two days ago, my kid asked me to figure out why an
    electric blanket
    wasn't working. The controller refused to turn on. I
    opened it and
    confirmed that it was getting supply voltage. Beyond
    that, the pile of
    dozens of surface mount electronic components was
    incomprehensible to
    me. I suspect Jeff might have been able to diagnose it,
    but not me.
    Here's a photo:
    https://www.flickr.com/photos/16972296@N08/54259119364/
    in/dateposted-public/
    I know electric blankets are old technology. I know they
    functioned well
    for decades with maybe a rheostat and perhaps a couple of
    other
    components. Why add unrepairable complexity?

    Don't know.  We use devices like the blue one in the
    following picture.
    <https://upload.wikimedia.org/wikipedia/commons/4/4b/
    W%C3%A4rmflasche1.jpg>
    and better isolating blankets. No electricity necessary. :-)

    :-) But you imply that _I'm_ the one insufficiently modern?

    A few lights in our house are switched by set of 2 x 3
    inexpensive
    wireless sockets including two remote controls, that I
    bought eleven
    years ago. I've still to replace the batteries. Two of the
    sockets are
    still spares, I have a replacement cell for the remote
    controls stored
    which might live even longer. Standard type, used in
    garage openers and
    burglar alarms, too. Selecting a channel and paring one of
    the four
    buttons of a remote control with one or more of the
    sockets is as easy
    as pie, using a line of dip switches inside those
    devices.  Quite
    similar to pairing switches and derailleur on our bicyles.

    While I avoid having essential functionality in my house
    depend on
    wireless connections, I enjoy having the option, for
    certain use cases,
    though.

    Yuck.

    I bought a new pedestal fan for our back patio last year.
    I've found a fan to be the best defense against mosquitos.
    The previous one still runs, but exposure to sunlight
    destroyed its plastic finger protection screen.

    Anyway, the new one comes with a remote control, roughly the
    size of a snack cracker. Who needs a remote to adjust a fan?
    I rejected ceiling fans with remotes when I was shopping,
    too. It's something to lose, something whose battery will
    die, and something whose electronics will go bad and be
    unrepairable.
    ...

    When I see a 50 year old, perfect looking bicycle, I think:
    That one must have been standing around inside most of the
    time, perhaps
    for a reason.

    My favorite bike is a young 39 years. It looks perfect, in
    part because I had it (and my wife's matching one) stripped
    and powder coated a couple years ago. Too many excellent
    memories to let it go.



    Not only the electronics infestation of consumer products.

    I brought food to make girlfriend's birthday dinner last
    evening and I cook in her kitchen regularly but sometimes
    the modern kitchen stove is a real impediment.

    [break for electronics rant: Her new kitchen stove, as her
    furnace, both run from the large LP tank in the yard. Both
    'improved' designs cannot run without electronic
    start/valves etc so when the power is out there is no heat
    whatsoever and power outages in rural USA are periodic.]

    The new kitchen stove has electronic controls for the
    burners and oven (oven controlled by a touch pad not, a
    physical switch). The burner controls have 4 positions
    only, viz., Off, Low, High, Start. That's really difficult
    for some cooking projects. Resolved by moving the two pans
    off and on the flame, which is Just One More Thing when
    timing two dishes at once.

    Only a designer who has never actually cooked would think of
    that. Works fine for brewing coffee or boiling pasta but
    limited for many projects.

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to zen cycle on Sun Jan 12 11:30:44 2025
    On 1/12/2025 6:22 AM, zen cycle wrote:
    On 1/11/2025 4:22 PM, Wolfgang Strobl wrote:
    Am Sat, 11 Jan 2025 00:06:46 -0500 schrieb zen cycle
    <funkmasterxx@hotmail.com>:


    I'm no fan of electronic shifting - it would be just my
    luck to be on
    planed 4-6 hour ride, and have something crap out at the
    furthest point
    from the car/house. That said, I'm never going back to
    discrete shifters.

    It isn't that difficult to carry a CR2032 and a tiny 25 g
    rechargeable
    battery somewhere. The coin cell lasts about two years,
    the battery is
    good for about 800-1000 km. Both warn early via LED on the
    device or via
    a Garmin Edge computer, for not having to care during a
    4-6 hour ride.


    I can't tell you how many times I leave the house and forget
    my wallet. Adding trying to remember to carry a spare
    battery is only part of the problem though.

    The other issue is the failure rates associated with a
    complex electro-mechanical systems VS a simple cable
    derailleur system. (Since I conduct FMEDAs as part of my
    role, I'm well-aware of the exponential failure rates
    associated with increased complexity).

    Since I've been seriously riding in the early 1980's, I've
    never had a problem associated with a shifting mechanism
    that I couldn't handle with a simple road/trail side repair,
    except for the few times when I've sheared the derailleur
    off in the woods. If your E-shifting mechanism fails for any
    reason other than the battery going dead, you're pretty much
    stick in whenever gear it left you in.

    I don't begrudge anyone for wanting E-shifting. I think it's
    a great technological advancement. The few times I've ridden
    E-shifting systems, I've been impressed with the accuracy,
    repeatability, and quickness of the shifts. But I don't see
    those advantages of being worth the extra cost and risk of
    failure. That's just my opinion, YMMV.

    Modern electronic shift systems work. They're different from
    mechanical but IMHO neither better nor worse, just different.

    https://pezcyclingnews.com/technspec/how-i-survived-a-dead-sram-etap-battery/

    https://www.reddit.com/r/cycling/comments/p0vxtq/shimano_di2_options_when_battery_dies_mid_trip/

    "One temporary solution would be to find someone that has a
    Di2, unplug their and your rear derailleur and plug in their
    cable into your derailleur. You can then change to the
    comfortable gear and ride back home. "

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From zen cycle@21:1/5 to AMuzi on Sun Jan 12 12:55:26 2025
    On 1/12/2025 12:30 PM, AMuzi wrote:
    On 1/12/2025 6:22 AM, zen cycle wrote:
    On 1/11/2025 4:22 PM, Wolfgang Strobl wrote:
    Am Sat, 11 Jan 2025 00:06:46 -0500 schrieb zen cycle
    <funkmasterxx@hotmail.com>:


    I'm no fan of electronic shifting - it would be just my luck to be on
    planed 4-6 hour ride, and have something crap out at the furthest point >>>> from the car/house. That said, I'm never going back to discrete
    shifters.

    It isn't that difficult to carry a CR2032 and a tiny 25 g rechargeable
    battery somewhere. The coin cell lasts about two years, the battery is
    good for about 800-1000 km. Both warn early via LED on the device or via >>> a Garmin Edge computer, for not having to care during a 4-6 hour ride.


    I can't tell you how many times I leave the house and forget my
    wallet. Adding trying to remember to carry a spare battery is only
    part of the problem though.

    The other issue is the failure rates associated with a complex
    electro-mechanical systems VS a simple cable derailleur system. (Since
    I conduct FMEDAs as part of my role, I'm well-aware of the exponential
    failure rates associated with increased complexity).

    Since I've been seriously riding in the early 1980's, I've never had a
    problem associated with a shifting mechanism that I couldn't handle
    with a simple road/trail side repair, except for the few times when
    I've sheared the derailleur off in the woods. If your E-shifting
    mechanism fails for any reason other than the battery going dead,
    you're pretty much stick in whenever gear it left you in.

    I don't begrudge anyone for wanting E-shifting. I think it's a great
    technological advancement. The few times I've ridden E-shifting
    systems, I've been impressed with the accuracy, repeatability, and
    quickness of the shifts. But I don't see those advantages of being
    worth the extra cost and risk of failure. That's just my opinion, YMMV.

    Modern electronic shift systems work. They're different from mechanical
    but IMHO neither better nor worse, just different.

    Which is why I wrote "The few times I've ridden E-shifting systems, I've
    been impressed with the accuracy, repeatability, and quickness of the
    shifts. "


    https://pezcyclingnews.com/technspec/how-i-survived-a-dead-sram-etap- battery/

    https://www.reddit.com/r/cycling/comments/p0vxtq/ shimano_di2_options_when_battery_dies_mid_trip/

    "One temporary solution would be to find someone that has a Di2, unplug
    their and your rear derailleur and plug in their cable into your
    derailleur. You can then change to the comfortable gear and ride back
    home. "


    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Roger Merriman@21:1/5 to Frank Krygowski on Sun Jan 12 18:36:56 2025
    Frank Krygowski <frkrygow@sbcglobal.net> wrote:
    On 1/11/2025 4:22 PM, Wolfgang Strobl wrote:
    Am Sat, 11 Jan 2025 00:06:46 -0500 schrieb zen cycle
    <funkmasterxx@hotmail.com>:


    I'm no fan of electronic shifting - it would be just my luck to be on
    planed 4-6 hour ride, and have something crap out at the furthest point
    from the car/house. That said, I'm never going back to discrete shifters. >>
    It isn't that difficult to carry a CR2032 and a tiny 25 g rechargeable
    battery somewhere. The coin cell lasts about two years, the battery is
    good for about 800-1000 km. Both warn early via LED on the device or via
    a Garmin Edge computer, for not having to care during a 4-6 hour ride.


    But how many threads have we had here about problems with Garmin computers?


    99.9% are from one poster…

    It’s a Him thing, Garmin/Wahoo/Hammerhead stuff is fairly mature tech now
    and on the whole just works.

    I have the same unit as him and unsurprisingly it hasn’t caused me any problems!

    Roger Merriman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Wolfgang Strobl@21:1/5 to All on Sun Jan 12 19:49:40 2025
    Am Sun, 12 Jan 2025 07:22:01 -0500 schrieb zen cycle <funkmasterxx@hotmail.com>:

    On 1/11/2025 4:22 PM, Wolfgang Strobl wrote:
    Am Sat, 11 Jan 2025 00:06:46 -0500 schrieb zen cycle
    <funkmasterxx@hotmail.com>:


    I'm no fan of electronic shifting - it would be just my luck to be on
    planed 4-6 hour ride, and have something crap out at the furthest point >>>from the car/house. That said, I'm never going back to discrete shifters.

    It isn't that difficult to carry a CR2032 and a tiny 25 g rechargeable
    battery somewhere. The coin cell lasts about two years, the battery is
    good for about 800-1000 km. Both warn early via LED on the device or via
    a Garmin Edge computer, for not having to care during a 4-6 hour ride.


    I can't tell you how many times I leave the house and forget my wallet. >Adding trying to remember to carry a spare battery is only part of the >problem though.

    This is a bit like comparing apples and oranges, though. After having
    left my house key behind once, I learned my lesson, I just check having
    the correct key, before closing the door, every time.

    I don't check batteries, when doing a day trip, simply because

    - the Edge 1030 does that for me
    - I would recharge the rechargeable batteries, before doing a multi
    day trip
    - I can't forget the batteries, because I carry spares in a bag that
    stays on the bike.
    - I carry a phone and enough money on my local trips in order to be
    able to call a taxi to take me and the bike home.

    See <https://www.mystrobl.de/ws/pic/fahrrad/20230609/P1068954.jpg>,
    the bag on the top tube is screwed to the top tube.

    So there isn't really anything special to remember. I carry some
    minature tools, too, a TPU tube, just in case. It is a sortiment that
    gets adjusted depending on circumstances, but rarely. YMMV.

    Checking and carrying one one of my cameras is more of a hassle. The
    SD-Card has to be emptied, sometimes. The batterie has to be checked and charged. But I carry a spare one for the camera, too.



    The other issue is the failure rates associated with a complex >electro-mechanical systems VS a simple cable derailleur system. (Since I >conduct FMEDAs as part of my role, I'm well-aware of the exponential
    failure rates associated with increased complexity).

    Problem is, you don't have reliable and useable statistics, neither for
    cables nor for wireless shifters. So this is still just guesswork.
    Admiring the simple construction of a Bowden cable won't help you if
    you're stranded far out with a broken cable and don't have a
    replacement. Or if the highly complex inner workings of a mechanical
    gearshift handle that is no longer sold have become defective. I
    experienced both, in the past.



    Since I've been seriously riding in the early 1980's, I've never had a >problem associated with a shifting mechanism that I couldn't handle with
    a simple road/trail side repair, except for the few times when I've
    sheared the derailleur off in the woods.

    Actually, my wife would have stranded last year somewhere in the middle
    of nowhere, if we wouldn't have done the tour together. So I cycled the
    ten or twenty kilometers back to our accommodation, got into the car,
    drove back and loaded the bike into the trunk.

    What happened? Well, the saddle clamp broke. As it turned out, the clamp included by the frame supplier for her bike was a cheap lemmon. It is
    almost impossible to ride up a steep hill without a saddle. The clamp on
    my bike was fine. But of course, I replaced both clamps, before riding
    again. I was quite annoyed that I didn't inspect the seat clamps more
    closely when I built the bikes. Lesson learned ...

    THe saddle clamp is almost the simplest mechanical part of a bike that I
    can think of.


    If your E-shifting mechanism
    fails for any reason other than the battery going dead, you're pretty
    much stick in whenever gear it left you in.

    Not true. It is possible to shift in both directions by pushing a tiny
    button on the derailleur. Good enough to get home. Doing something
    similar with a broken Bowden cable is possible, but difficult. I doubt
    that your average cyclist can do it.

    It happended with the purely mechanical shifter on my former bicycle,
    that part broke beyond repair, without warning, during the last ride
    before our yearly vacation. Just by luck, I found a dealer who still had
    a replacement in stock, so that I got the bike back one day before
    departure.

    I wouldn't use consider wireless shifting on a bike intended for a trip
    around the world or through a developing country. But I wouldn't use
    most of what has been sold during the past three decades as a purely
    mechanical system, either.


    I don't begrudge anyone for wanting E-shifting. I think it's a great >technological advancement. The few times I've ridden E-shifting systems,
    I've been impressed with the accuracy, repeatability, and quickness of
    the shifts. But I don't see those advantages of being worth the extra
    cost and risk of failure.

    Extra cost is a thing, indeed. But that will change. Small actuators,
    tiny CPUs and ANT+/BLE variants aren't rocket science or really new,
    anymore. IMO, a reliable E-shifting system like those currently sold by
    Shimano and SRAM could be built to be both more reliable and cheaper
    than their older mechanical counterparts (say Shimano R8000). Fewer
    mechanical parts, even fewer electronic parts with a known MTBF would
    allow that. Given that we a surrounded by long lived electronic devices
    that are far more complex than these very simple E-Shifting systems, it
    needs really bad luck or bad intentions to make them short-lived.
    Actually, it is a wonder that modern cars don't break shortly after
    leaving the factory. :-) And of course, I don't know how reliable those electronic parts from SRAM and Shimano really are. But I know that some mechanical shifters from Shimano are less reliable than I like.


    That's just my opinion, YMMV.

    Sure. :-) So far, most of this is just guesswork. Wait and see ...


    --
    Thank you for observing all safety precautions

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to funkmasterxx@hotmail.com on Sun Jan 12 11:03:13 2025
    On Sun, 12 Jan 2025 07:22:01 -0500, zen cycle
    <funkmasterxx@hotmail.com> wrote:

    I can't tell you how many times I leave the house and forget my wallet. >Adding trying to remember to carry a spare battery is only part of the >problem though.

    I had the same problem. My solution was to put a cardboard box near
    the door that contains everything I plan to take with me for the day,
    including my wallet and TWO car keys. When I leave the house, I only
    need to remember to either fill my pockets with the contents of the
    box or just take the box with me. If I don't do that, trying to
    remember what I need to take with me in the morning is failure prone
    (if not futile). I've been using this system successfully for about
    30 years. The only times this system fails is when I do something
    impulsive.

    I have a similar cardboard box in my car for stuff that needs to be
    dragged into the house. When I still had my former palatial office, I
    also put a box near the office door.

    If you're planning on going for a ride, you probably will have done
    some planning the previous day. Instead of trying to remember to
    bring a spare battery, just throw it in the box.


    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Frank Krygowski on Sun Jan 12 13:45:11 2025
    On 1/12/2025 1:25 PM, Frank Krygowski wrote:
    On 1/12/2025 7:49 AM, zen cycle wrote:
    On 1/11/2025 7:46 PM, Frank Krygowski wrote:

    But individual power outputs from left and right legs? I
    can't imagine much value from that measurement.

    You answered your own question. "You" can't imagine it's
    useful. There are a great many data geeks who find it
    quite useful.

    Maybe we can take a poll? How many here gather and use data
    on their individual legs' power output?


    If he finds it interesting, he ought to pursue it, even to
    recording and manipulating the data (daily/weekly/monthly
    averages, min-max differential and the like).

    If you, as I, have no interest whatsoever, we shall not
    acquire one.

    As with so many discussions here, personal taste, for any
    reason or for no reason, is reason enough for a decision
    either way.

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Catrike Ryder@21:1/5 to frkrygow@sbcglobal.net on Sun Jan 12 15:00:04 2025
    On Sun, 12 Jan 2025 14:25:37 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/12/2025 7:49 AM, zen cycle wrote:
    On 1/11/2025 7:46 PM, Frank Krygowski wrote:

    But individual power outputs from left and right legs? I can't imagine
    much value from that measurement.

    You answered your own question. "You" can't imagine it's useful. There
    are a great many data geeks who find it quite useful.

    Maybe we can take a poll? How many here gather and use data on their >individual legs' power output?


    What would a poll of a bunch of old RBT geezers prove?

    --
    C'est bon
    Soloman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Frank Krygowski on Sun Jan 12 14:24:14 2025
    On 1/12/2025 2:12 PM, Frank Krygowski wrote:
    On 1/12/2025 2:45 PM, AMuzi wrote:
    On 1/12/2025 1:25 PM, Frank Krygowski wrote:
    On 1/12/2025 7:49 AM, zen cycle wrote:
    On 1/11/2025 7:46 PM, Frank Krygowski wrote:

    But individual power outputs from left and right legs?
    I can't imagine much value from that measurement.

    You answered your own question. "You" can't imagine it's
    useful. There are a great many data geeks who find it
    quite useful.

    Maybe we can take a poll? How many here gather and use
    data on their individual legs' power output?


    If he finds it interesting, he ought to pursue it, even to
    recording and manipulating the data (daily/weekly/monthly
    averages, min-max differential and the like).

    If you, as I, have no interest whatsoever, we shall not
    acquire one.

    As with so many discussions here, personal taste, for any
    reason or for no reason, is reason enough for a decision
    either way.

    That's fine. Choice is allowed. I was just interested in
    actual quantities of "a great many data geeks who find it
    quite useful" using our present sample.

    I don't think it's unusual to have zero interest in
    individual leg power.

    YMMV, of course.



    Right, not for everyone.
    I think the stipulation, "data geeks" was clear!

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Catrike Ryder@21:1/5 to frkrygow@sbcglobal.net on Sun Jan 12 15:26:48 2025
    On Sun, 12 Jan 2025 15:12:38 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/12/2025 2:45 PM, AMuzi wrote:
    On 1/12/2025 1:25 PM, Frank Krygowski wrote:
    On 1/12/2025 7:49 AM, zen cycle wrote:
    On 1/11/2025 7:46 PM, Frank Krygowski wrote:

    But individual power outputs from left and right legs? I can't
    imagine much value from that measurement.

    You answered your own question. "You" can't imagine it's useful.
    There are a great many data geeks who find it quite useful.

    Maybe we can take a poll? How many here gather and use data on their
    individual legs' power output?


    If he finds it interesting, he ought to pursue it, even to recording and
    manipulating the data (daily/weekly/monthly averages, min-max
    differential and the like).

    If you, as I, have no interest whatsoever, we shall not acquire one.

    As with so many discussions here, personal taste, for any reason or for
    no reason, is reason enough for a decision either way.

    That's fine. Choice is allowed. I was just interested in actual
    quantities of "a great many data geeks who find it quite useful" using
    our present sample.

    I don't think it's unusual to have zero interest in individual leg power.

    YMMV, of course.

    But then, I'll bet you don't even think it's useful to keep track of
    your cadence.

    --
    C'est bon
    Soloman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Wolfgang Strobl@21:1/5 to All on Sun Jan 12 21:45:36 2025
    Am Sun, 12 Jan 2025 11:30:44 -0600 schrieb AMuzi <am@yellowjersey.org>:

    Modern electronic shift systems work. They're different from
    mechanical but IMHO neither better nor worse, just different.

    https://pezcyclingnews.com/technspec/how-i-survived-a-dead-sram-etap-battery/

    Didn't have to read that in order to know that he just installed the
    battery of the front derailleur on the back derailleur. One of our kids,
    who bought a wireless SRAM group before I built our bikes, came up with
    that idea during a ride, himself. Necessity is the mother of invention,
    as they say. But like us, he now has spare battery and a CR2032, in
    some of his pockets. A single energy bar needs more space.


    --
    Thank you for observing all safety precautions

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Wolfgang Strobl@21:1/5 to All on Sun Jan 12 21:33:05 2025
    Am Sat, 11 Jan 2025 19:46:50 -0500 schrieb Frank Krygowski <frkrygow@sbcglobal.net>:

    On 1/11/2025 4:11 PM, Wolfgang Strobl wrote:
    Am Fri, 10 Jan 2025 20:01:26 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    On 1/10/2025 5:18 PM, Wolfgang Strobl wrote:

    This makes me reflect on the criticism of electronics in bicycles. ... But what about measuring the
    amount of power applied to the pedals, what about telling the cyclist
    who balanced he splits the power between left and right, by
    instrumenting the pedal or the bottom bracket?

    Yes, that can be done with enough sensors and electronics. But it seems
    like useless information to me. As with much computer software, it seems >>> like "feature bloat." Why would anybody but a racer care?

    Is there any reason to believe that old people like me don't have a need
    to care about training intensity, as a matter of principle? How comes?

    When using my biycles as a middle aged, healthy adult I didn't care
    about racing, didn't train and didn't try to find company in cycling
    clubs, either. I didn't visit a gym until much later. Cycling was simply
    a way to get to work quickly. During vacations it was a more
    entertaining mode of transportation than the car. Fitness, strength and
    cardio fitness came as a side effect, slowly and almost unnoticed.

    This has changed. My cardio fitness is still better than what I know
    about a lot of people half my age. Nevertheless, it decreased over the
    years and it takes more and more effort to keep what you have not yet
    lost. As it is expected. The ability to measure power is helpful in
    finding the point at which intensity is sufficient and overload has not
    yet begun. This is even more true in combination with a heart rate
    monitor.

    So far, this is only about aging. But there is more.

    Accidents that led to damage to bones, joints and tendons in the past
    have consequences to be considered, too. There is a tendency to
    compensate weaknesses by bad postures, whithout noticing. This is both
    caused by damages and causing damages. This is to be avoided. Measuring
    how power is applied by the feet helps detecting differences early,
    avoiding damage.

    Avoiding damage is far from useless.

    But individual power outputs from left and right legs? I can't imagine
    much value from that measurement.

    So imagine somebody who measured 49%/51% +-1% most of the time, during a
    few years in the past. Can't you imagine a reason why this could have
    changed to 47%/53%, after a ride? What about a sudden change to 55%/45%
    or worse, after getting back to 50%, 50%?

    In fact, I don't believe power
    measurement is necessary for fitness.

    "Necessary" has to do quite some heavy lifting, in this sentence. :-)

    <https://www.ipmultisport.com/ref_lib/Coggan_Power_Meter.pdf>


    I judge by feel. It has the
    advantage of naturally compensating for days when I feel stronger or
    weaker. And it's organic! ;-)

    My orthopedist told me, "that training will be painfull, but it is
    necessary". No joking.



    My experiences with mechanical devices are at least as mixed as those
    with electronic devices. Some very complicated electronic devices that
    I bought decades ago still work. The same applies to some that I built
    myself a long time ago. With devices that contained both electronic and
    mechanical parts, the fault was more often on the mechanical side.

    Anecdotical, I know.

    To me, a big advantage is the ability to _look_ at a mechanical device
    and _see_ what's wrong.

    That's professional blindness. Leaving aside the tools and materials
    used by a blacksmith, mechanical devices and materials often have
    defects that cannot be seen, but have to be measured or looked up. Same
    with electronics. Judging an electronic component - say, an 50 cent microcontroller/FGPA/ASIC - by appearance or the number of gates is
    misleading.

    That, and the fact I can often affect a repair.

    I prefer devices that don't need repair over their lifetime. Of course, answering the question "which is which?" it is often just guesswork. I
    just doubt that mechatronics is fundamentally less reliable than
    mechanics from the days of the cuckoo clock.


    I hate the Kleenex ethic - "It's no good any more, just throw it away."

    Sometimes I reuse a 0.5 cent screw, even when it is somewhat damaged.
    For example when a replacement isn't easy to get by. But I avoid
    risking an expensive device that I depend on by desoldering an old elko
    from an old radio in order to save a few cent. Same with electronic
    components that are tiny and cheap and come with known specifications.



    Two days ago, my kid asked me to figure out why an electric blanket
    wasn't working. The controller refused to turn on. I opened it and
    confirmed that it was getting supply voltage. Beyond that, the pile of
    dozens of surface mount electronic components was incomprehensible to
    me. I suspect Jeff might have been able to diagnose it, but not me.
    Here's a photo:
    https://www.flickr.com/photos/16972296@N08/54259119364/in/dateposted-public/
    I know electric blankets are old technology. I know they functioned well >>> for decades with maybe a rheostat and perhaps a couple of other
    components. Why add unrepairable complexity?

    Don't know. We use devices like the blue one in the following picture.
    <https://upload.wikimedia.org/wikipedia/commons/4/4b/W%C3%A4rmflasche1.jpg> >> and better isolating blankets. No electricity necessary. :-)

    :-) But you imply that _I'm_ the one insufficiently modern?

    No. I doubt that "being modern" is a reasonable benchmark or measure.
    This cuts both ways. :-)



    A few lights in our house are switched by set of 2 x 3 inexpensive
    wireless sockets including two remote controls, that I bought eleven
    years ago. I've still to replace the batteries. Two of the sockets are
    still spares, I have a replacement cell for the remote controls stored
    which might live even longer. Standard type, used in garage openers and
    burglar alarms, too. Selecting a channel and paring one of the four
    buttons of a remote control with one or more of the sockets is as easy
    as pie, using a line of dip switches inside those devices. Quite
    similar to pairing switches and derailleur on our bicyles.

    While I avoid having essential functionality in my house depend on
    wireless connections, I enjoy having the option, for certain use cases,
    though.

    Yuck.

    You don't have a single remote control in your house? Not even for the
    TV? That's rare.



    I bought a new pedestal fan for our back patio last year. I've found a
    fan to be the best defense against mosquitos. The previous one still
    runs, but exposure to sunlight destroyed its plastic finger protection >screen.

    Anyway, the new one comes with a remote control, roughly the size of a
    snack cracker. Who needs a remote to adjust a fan? I rejected ceiling
    fans with remotes when I was shopping, too. It's something to lose,
    something whose battery will die, and something whose electronics will
    go bad and be unrepairable.

    Strange. Microelectronics using the smallest available microcontrollers
    that still are breadboard compatible has been one of my hobbies over the
    years. I still have a box of old, but still working IR remote controls,
    from devices that broke many years ago. I collected these to control
    gadgets like this one <https://www.mystrobl.de/Plone/basteleien/microcontroller/ws2812/DSC_3564-DSC_3566_fused.JPG>
    Can you guess what this blinkenlight does?

    Parts a an IR receiver, a PIC 12F1840 microcontroller, a stripe of eight
    RGB LED, and a remote control from a CD Player that broke long before
    2014, when I built that gadget. The aforementioned $1.50 controller
    (single quantity, digikey) does everything from IR decoding to
    controlling the LED stripe.

    Somewhat later, I built something larger using a different part (an
    ESP8266) for illuminating the house bar of one of our kids, using about
    one meter of densely placed RGB LED, controlled via WIFI, doing a whole
    series of different colorful light effects. Extendable with new effects
    by uploading short LUA snippets, of course. :-) I was told that it was
    used again at a New Year's celebration, so obviously it is still
    working.

    That somewhat outdated ESP8622 is still available via Amazon or Alibaba,
    for between $0.50 and $2, depending on whether you want the bare bones
    platine or a development board. Mass produced electronics are cheap,
    nowadays.


    ...

    When I see a 50 year old, perfect looking bicycle, I think:
    That one must have been standing around inside most of the time, perhaps
    for a reason.

    My favorite bike is a young 39 years. It looks perfect, in part because
    I had it (and my wife's matching one) stripped and powder coated a
    couple years ago. Too many excellent memories to let it go.

    My wife still uses an almost as old bicycle for everyday rides around
    the corner. Didn't have to strip and power coat it, because it came
    that way, when she bought it.

    But like me, she is glad that I build two road bikes in 2023, using
    wireless electronic shifting that you dislike so much. Without, she
    wouldn't have been able to do some of those very enjoyable tours
    throughout the region that we did in 2023 and in 2024. Neither would I,
    for most of the longer and faster tours that I did riding alone, during
    that time.

    Before you ask, certainly I could have bought the somewhat cheaper
    mechanical parts. But given that at least two of those tours that I
    enjoyed and valued were at my limit, given that installing and adjusting
    the mechanical version would have been more work, I'm happy about about
    my choice.


    --
    Bicycle helmets are the Bach flower remedies of traffic

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to frkrygow@sbcglobal.net on Sun Jan 12 14:13:36 2025
    On Fri, 10 Jan 2025 20:01:26 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    Two days ago, my kid asked me to figure out why an electric blanket
    wasn't working. The controller refused to turn on. I opened it and
    confirmed that it was getting supply voltage. Beyond that, the pile of
    dozens of surface mount electronic components was incomprehensible to
    me. I suspect Jeff might have been able to diagnose it, but not me.
    Here's a photo: >https://www.flickr.com/photos/16972296@N08/54259119364/in/dateposted-public/

    I don't think I can do much with just a photograph. If you have the
    maker, model and FCC ID number, I can do better. Also, some
    indication of what it does when the power is applied, such as do the
    LED's light. If the don't light, try again in the dark and see if
    they partially light.

    Looking at the PCB components, I would guess it's 2000 to 2010
    vintage. In other words, it's old. I can't read any of the part
    numbers from the photo. If the SOT-24 package (probably a PIC
    Microcontroller) has a date, that would determine the age. Looking at
    an enlargement of the photo, I notice that some of the components
    (LED's and pushbutton switches) were hand soldered. Enlarging the
    photo showed several possible places where the soldering looks like a
    volcano with a black annular ring around the center component lead.
    That's about all I can do with just a photo.

    I know electric blankets are old technology. I know they functioned well
    for decades with maybe a rheostat and perhaps a couple of other
    components. Why add unrepairable complexity?

    I don't want to write a lengthy essay on topic. Your rheostat
    probably was fine for mid 20th century technology, but would never
    pass the safety standards that followed. Those safety standards were
    there for a good reason such as killing the customer by fire or
    electrocution. The rheostat design has exactly one feature, which is
    to manually adjust the temperature. If you want more features
    (overload disconnect, timer, twin bed control, fuzzy logic control,
    etc) electronics is required.

    Incidentally, I've had several occasions where customers demanded
    their old and reliable computer back, usually because they were
    allergic to reading and following the instructions. Eventually, all
    of them went back to the latest greatest electronic model because the
    old and reliable one was slow, difficult to use, or didn't do
    something they needed. That's the real problem with technology. It's
    really difficult to go back to primitive.

    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to news51@mystrobl.de on Sun Jan 12 15:23:33 2025
    On Fri, 10 Jan 2025 23:18:27 +0100, Wolfgang Strobl
    <news51@mystrobl.de> wrote:

    Am Fri, 10 Jan 2025 07:25:48 -0800 schrieb Jeff Liebermann ><jeffl@cruzio.com>:
    Humans tend to prefer whatever technology they learned first.

    Not really. I learned programming by creating punched cards using an IBM
    26 Printing Card Punch, editing those programs by using the card
    duplicating feature of those machines. Followed by pinnig stretches of
    pages printed by large chain printers like to a wall, in order to study >complaints from the Fortran IV compiler, or using a pen to mark logical >errors in the source code, for later correction work on said card punch.

    <https://en.wikipedia.org/wiki/IBM_1403>

    I started with the IBM 1620 running Gotran (a small version of Fortran
    II) and SPS for assembly language.
    <https://en.wikipedia.org/wiki/IBM_1620> <https://archive.org/details/bitsavers_ibm1620C26nterpretiveProgrammingSystem1961_2419558>
    <https://bitsavers.trailing-edge.com/pdf/ibm/1620/monitor/C26-5793-0_1620_SPS_III_Programming_System_Reference_1964.pdf>
    The 1620 was later resurrected in the Forbin Project movie as
    Colossus. I was hooked on computing (hardware) after I discovered the
    well hidden secret drawer under the console desk where the admin's hid
    all the manuals and paperwork.

    My
    first calculator was basically a mechanical "adding machine". >><https://www.google.com/search?num=10&q=marchant+adding+machine&udm=2>
    I eventually ended up with an HP-35 RPN calculator and loved it. RPN
    is easy, if you think like a computer that stores intermediate results
    in a stack.

    Of course. If you do.

    I do. The problem was that we didn't really have computers with
    stacks until the IBM 360.

    There is nothing wrong with RPN, if it works for you. But there isn't >anything "natural" about it.

    It was "natural" because it was easier for me to use.

    I rarely used calculators, because writing short program snippets is
    more natural for me. Many decades ago I did that using PL/I-80 on my >self-built CP/M computer, nowadays I just fire up a Python REPL, or
    start a Jupyter notebook, for so called "back of the envelope"
    calculations.

    I'm lazy. For anything complicated or highly repetitive, I use Google
    search to find a web app or dedicated calculator for whatever I'm
    trying to calculate. It will probably be written in Javascript. I
    can usually find a suitable online calculator for almost all the
    electronic calculations I might need. If I need something obscure or
    can't find anything, I'll scribble something for the occasion. I'm
    currently dabbling with AI code generators. I'm a rather bad
    programmer, so learning will take some time.

    My first and only HP calculator is a HP 200LX, still working fine.

    I had an HP95, HP 100LX, and a HP200LX. I still have the HP100-LX,
    but the others were sold long ago. Trying to find PCMCIA cards with
    small amounts of memory, is what forced me to stop using them. They
    were really "cool" in their day.

    I
    lost one of the tiny case screws, decades ago, replacing the CR2032 now
    and then is still awkward. I have no real use for the device anymore,
    but I still like it enough for not giving it away. AFAIR, there's still
    a copy of the original DeSmet C compiler on the flash card in the PCMCIA >slot. :-)

    Ok, you're addicted. I have a rather disorganized box of HP
    calculator parts. I'll look for a screw and mail it to you.
    Otherwise, there seem to be some available online: <https://www.200lx.net/mtncekit.htm>

    I removed one screw from my HP 100LX to check the size. I would
    guess(tm) it's 2-56, but without a gauge or a 2-56 nut, I can't be
    sure. However, I can't find my thread gauges and my microscope
    threading graticule is in use. So this will need to wait. Please
    remind me if I forget.

    Getting back to cycling ... :-)

    Sorry. I need to get back to house cleaning and firewood hauling.
    I'll read through the rest of your posting and write something,
    hopefully in a few days.

    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to John B. on Sun Jan 12 19:28:34 2025
    On 1/12/2025 7:19 PM, John B. wrote:
    On Sun, 12 Jan 2025 11:08:45 -0600, AMuzi <am@yellowjersey.org> wrote:

    On 1/11/2025 6:46 PM, Frank Krygowski wrote:
    On 1/11/2025 4:11 PM, Wolfgang Strobl wrote:
    Am Fri, 10 Jan 2025 20:01:26 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    On 1/10/2025 5:18 PM, Wolfgang Strobl wrote:
    This makes me reflect on the criticism of electronics in
    bicycles. ... But what about measuring the
    amount of power applied to the pedals, what about
    telling the cyclist
    who balanced he splits the power between left and right, by
    instrumenting the pedal or the bottom bracket?

    Yes, that can be done with enough sensors and
    electronics. But it seems
    like useless information to me. As with much computer
    software, it seems
    like "feature bloat." Why would anybody but a racer care?

    Is there any reason to believe that old people like me
    don't have a need
    to care about training intensity, as a matter of
    principle?  How comes?

    When using my biycles as a middle aged, healthy adult I
    didn't care
    about racing, didn't train and didn't try to find company
    in cycling
    clubs, either. I didn't visit a gym until much later.
    Cycling was simply
    a way to get to work quickly. During vacations it was a more
    entertaining mode of transportation than the car. Fitness,
    strength and
    cardio fitness came as a side effect, slowly and almost
    unnoticed.

    This has changed. My cardio fitness is still better than
    what I know
    about a lot of people half my age. Nevertheless, it
    decreased over the
    years and it takes more and more effort to keep what you
    have not yet
    lost. As it is expected. The ability to measure power is
    helpful in
    finding the point at which intensity is sufficient and
    overload has not
    yet begun. This is even more true in combination with a
    heart rate
    monitor.

    So far, this is only about aging. But there is more.

    Accidents that led to damage to bones, joints and tendons
    in the past
    have consequences to be considered, too. There is a
    tendency to
    compensate weaknesses by bad postures, whithout noticing.
    This is both
    caused by damages and causing damages. This is to be
    avoided.  Measuring
    how power is applied by the feet helps detecting
    differences early,
    avoiding damage.

    Avoiding damage is far from useless.

    But individual power outputs from left and right legs? I
    can't imagine much value from that measurement. In fact, I
    don't believe power measurement is necessary for fitness. I
    judge by feel. It has the advantage of naturally
    compensating for days when I feel stronger or weaker. And
    it's organic!  ;-)

    My experiences with mechanical devices are at least as
    mixed as those
    with electronic devices.  Some very complicated electronic
    devices that
    I bought decades ago still work. The same applies to some
    that I built
    myself a long time ago.  With devices that contained both
    electronic and
    mechanical parts, the fault was more often on the
    mechanical side.

    Anecdotical, I know.

    To me, a big advantage is the ability to _look_ at a
    mechanical device and _see_ what's wrong. That, and the fact
    I can often affect a repair. I hate the Kleenex ethic -
    "It's no good any more, just throw it away."

    Two days ago, my kid asked me to figure out why an
    electric blanket
    wasn't working. The controller refused to turn on. I
    opened it and
    confirmed that it was getting supply voltage. Beyond
    that, the pile of
    dozens of surface mount electronic components was
    incomprehensible to
    me. I suspect Jeff might have been able to diagnose it,
    but not me.
    Here's a photo:
    https://www.flickr.com/photos/16972296@N08/54259119364/
    in/dateposted-public/
    I know electric blankets are old technology. I know they
    functioned well
    for decades with maybe a rheostat and perhaps a couple of
    other
    components. Why add unrepairable complexity?

    Don't know.  We use devices like the blue one in the
    following picture.
    <https://upload.wikimedia.org/wikipedia/commons/4/4b/
    W%C3%A4rmflasche1.jpg>
    and better isolating blankets. No electricity necessary. :-)

    :-) But you imply that _I'm_ the one insufficiently modern?

    A few lights in our house are switched by set of 2 x 3
    inexpensive
    wireless sockets including two remote controls, that I
    bought eleven
    years ago. I've still to replace the batteries. Two of the
    sockets are
    still spares, I have a replacement cell for the remote
    controls stored
    which might live even longer. Standard type, used in
    garage openers and
    burglar alarms, too. Selecting a channel and paring one of
    the four
    buttons of a remote control with one or more of the
    sockets is as easy
    as pie, using a line of dip switches inside those
    devices.  Quite
    similar to pairing switches and derailleur on our bicyles.

    While I avoid having essential functionality in my house
    depend on
    wireless connections, I enjoy having the option, for
    certain use cases,
    though.

    Yuck.

    I bought a new pedestal fan for our back patio last year.
    I've found a fan to be the best defense against mosquitos.
    The previous one still runs, but exposure to sunlight
    destroyed its plastic finger protection screen.

    Anyway, the new one comes with a remote control, roughly the
    size of a snack cracker. Who needs a remote to adjust a fan?
    I rejected ceiling fans with remotes when I was shopping,
    too. It's something to lose, something whose battery will
    die, and something whose electronics will go bad and be
    unrepairable.
    ...

    When I see a 50 year old, perfect looking bicycle, I think:
    That one must have been standing around inside most of the
    time, perhaps
    for a reason.

    My favorite bike is a young 39 years. It looks perfect, in
    part because I had it (and my wife's matching one) stripped
    and powder coated a couple years ago. Too many excellent
    memories to let it go.



    Not only the electronics infestation of consumer products.

    I brought food to make girlfriend's birthday dinner last
    evening and I cook in her kitchen regularly but sometimes
    the modern kitchen stove is a real impediment.

    [break for electronics rant: Her new kitchen stove, as her
    furnace, both run from the large LP tank in the yard. Both
    'improved' designs cannot run without electronic
    start/valves etc so when the power is out there is no heat
    whatsoever and power outages in rural USA are periodic.]

    The new kitchen stove has electronic controls for the
    burners and oven (oven controlled by a touch pad not, a
    physical switch). The burner controls have 4 positions
    only, viz., Off, Low, High, Start. That's really difficult
    for some cooking projects. Resolved by moving the two pans
    off and on the flame, which is Just One More Thing when
    timing two dishes at once.

    Only a designer who has never actually cooked would think of
    that. Works fine for brewing coffee or boiling pasta but
    limited for many projects.

    Gee,,, my mother cooked on a "gas" stove with mechanical controls from
    the regulate at the tank to the knobs on the stove successfully and
    one grandmother cooked on a wood stove, in the winter, un till she
    died, successfully.

    Are these "modern contraptions" really necessary?

    Well, perhaps they are. Are there young woman, today, being taught by
    their mother how to cook on a wood stove? Or how to darn woolen socks,
    or even how to raise the kids?


    Much agreed.

    The kitchen stove I used for decades had simple gas valves
    and a box of wooden matches in a clay dish next to it. No
    complaints.

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to frkrygow@sbcglobal.net on Sun Jan 12 19:33:00 2025
    On Sun, 12 Jan 2025 21:11:37 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/12/2025 5:13 PM, Jeff Liebermann wrote:
    On Fri, 10 Jan 2025 20:01:26 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    Two days ago, my kid asked me to figure out why an electric blanket
    wasn't working. The controller refused to turn on. I opened it and
    confirmed that it was getting supply voltage. Beyond that, the pile of
    dozens of surface mount electronic components was incomprehensible to
    me. I suspect Jeff might have been able to diagnose it, but not me.
    Here's a photo:
    https://www.flickr.com/photos/16972296@N08/54259119364/in/dateposted-public/

    I don't think I can do much with just a photograph. If you have the
    maker, model and FCC ID number, I can do better. Also, some
    indication of what it does when the power is applied, such as do the
    LED's light. If the don't light, try again in the dark and see if
    they partially light.

    There was no action at all, no lights, nothing.

    That makes troubleshooting easy. There has to be a fuse or thermal
    circuit breaker in there somewhere. I can't see one on the PCB
    (printed circuit board). It's a single sided PCB so it's unlikely
    that it's hiding on the back of the PCB. It's either inside the wall
    plug, inside some kind of on/off switch box that's in series with the
    power cord, or attached to the blanket somewhere.

    I took a 2nd look at the PCB and found some more things that look
    suspicious. <https://www.learnbydestroying.com/jeffl/crud/electric-blanket-control.jpg> It's a single sided (nothing but the cheapest) PCB. That means that
    any wires going through the pads on the PCB need to be well soldered
    or they will literally fall off. I circled 6 solder connections that
    don't look quite right. (Oops. I missed one). Compare these solder connections with the ones that look good on the left side of the PCB.

    There are also 4 connection on the upper left with don't look right.
    What is the black stuff on the leads? Melted plastic? The labels S1
    S2 are wires to a switch. H1 and H2 are heater wires.

    If you want me to continue, please provide the maker, model number,
    and FCC ID if available.

    Looking at the PCB components, I would guess it's 2000 to 2010
    vintage. In other words, it's old. I can't read any of the part
    numbers from the photo. If the SOT-24 package (probably a PIC
    Microcontroller) has a date, that would determine the age. Looking at
    an enlargement of the photo, I notice that some of the components
    (LED's and pushbutton switches) were hand soldered. Enlarging the
    photo showed several possible places where the soldering looks like a
    volcano with a black annular ring around the center component lead.
    That's about all I can do with just a photo.

    I thought about bad solder joints and looked pretty carefully, but
    didn't see any that looked suspicious.

    After I look for suspicious cold solder connections, I just resolder
    the connections.

    I didn't go so far as to follow
    the traces to see where voltage went away. The owner is quite prosperous
    and lives a distance away, and was saying "Dinner is ready!" She'll just
    buy a new one. I figure I did my duty by eliminating obvious faults and
    by pointing out that the blanket itself was still good, so just replace
    the controller.

    We have a different view of the problem. I look my duty is to prevent
    yet another piece of almost working electronics from ending up in the
    recycling bin.

    "Recycling Isnt the Answer; Its the Last Resort" <https://www.ifixit.com/Right-to-Repair/Recycling>

    Yes, I surmised that was the motivation. The terrible plague of electric >blanket deaths has finally been conquered! We are safer every year!

    Thank you for the cynical comment of the day.

    "Electric Blanket Fire Lawyer" <https://www.killinofirm.com/practice-areas/electric-blanket-fire-lawyer> "Electric blankets cause about 5,000 house fires per year in the US,
    resulting in dozens of injuries and even death."


    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to All on Sun Jan 12 23:01:12 2025
    On Mon, 13 Jan 2025 13:20:04 +0700, John B. <slocombjb@gmail.com>
    wrote:

    But long before the days of transistors and "circuit boards" my folks
    had an electric blanket... used it all the rest of their lives. Is the
    modern one better?

    Not really better, but certainly safer. Electric heaters are all 100% efficient. Every watt you shove into the heating wires is turned into
    heat. There's no way to improve on 100% efficiency.

    The problem with electric blankets is that they're rather fragile.
    Dogs, cats and rats chew on them causing exposed wires. Using them
    folded can break wires at the bend or cause overheating. There are
    probably other ways to damage the wires. Once exposed and/or broken,
    the safety question becomes what will the circuitry to prevent sparks, localized overheating, starting a fire, overload, etc. Dealing with
    all those potential problems doesn't create a better heating
    experience, but does make the blanket safer. Basically, you don't
    know that you have a problem until after the house burns down.

    Electric heated blanket recalls: <https://www.cpsc.gov/search?search_api_fulltext=electric+blankets>
    If you check the "Include archived items" box, you'll see all the old
    recalls.


    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Catrike Ryder@21:1/5 to All on Mon Jan 13 03:43:27 2025
    On Mon, 13 Jan 2025 13:20:04 +0700, John B. <slocombjb@gmail.com>
    wrote:

    On Sun, 12 Jan 2025 19:33:00 -0800, Jeff Liebermann <jeffl@cruzio.com>
    wrote:

    On Sun, 12 Jan 2025 21:11:37 -0500, Frank Krygowski >><frkrygow@sbcglobal.net> wrote:

    On 1/12/2025 5:13 PM, Jeff Liebermann wrote:
    On Fri, 10 Jan 2025 20:01:26 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    Two days ago, my kid asked me to figure out why an electric blanket
    wasn't working. The controller refused to turn on. I opened it and
    confirmed that it was getting supply voltage. Beyond that, the pile of >>>>> dozens of surface mount electronic components was incomprehensible to >>>>> me. I suspect Jeff might have been able to diagnose it, but not me.
    Here's a photo:
    https://www.flickr.com/photos/16972296@N08/54259119364/in/dateposted-public/

    I don't think I can do much with just a photograph. If you have the
    maker, model and FCC ID number, I can do better. Also, some
    indication of what it does when the power is applied, such as do the
    LED's light. If the don't light, try again in the dark and see if
    they partially light.

    There was no action at all, no lights, nothing.

    That makes troubleshooting easy. There has to be a fuse or thermal
    circuit breaker in there somewhere. I can't see one on the PCB
    (printed circuit board). It's a single sided PCB so it's unlikely
    that it's hiding on the back of the PCB. It's either inside the wall
    plug, inside some kind of on/off switch box that's in series with the
    power cord, or attached to the blanket somewhere.

    I took a 2nd look at the PCB and found some more things that look >>suspicious. >><https://www.learnbydestroying.com/jeffl/crud/electric-blanket-control.jpg> >>It's a single sided (nothing but the cheapest) PCB. That means that
    any wires going through the pads on the PCB need to be well soldered
    or they will literally fall off. I circled 6 solder connections that
    don't look quite right. (Oops. I missed one). Compare these solder >>connections with the ones that look good on the left side of the PCB.

    There are also 4 connection on the upper left with don't look right.
    What is the black stuff on the leads? Melted plastic? The labels S1
    S2 are wires to a switch. H1 and H2 are heater wires.

    If you want me to continue, please provide the maker, model number,
    and FCC ID if available.

    Looking at the PCB components, I would guess it's 2000 to 2010
    vintage. In other words, it's old. I can't read any of the part
    numbers from the photo. If the SOT-24 package (probably a PIC
    Microcontroller) has a date, that would determine the age. Looking at >>>> an enlargement of the photo, I notice that some of the components
    (LED's and pushbutton switches) were hand soldered. Enlarging the
    photo showed several possible places where the soldering looks like a
    volcano with a black annular ring around the center component lead.
    That's about all I can do with just a photo.

    I thought about bad solder joints and looked pretty carefully, but
    didn't see any that looked suspicious.

    After I look for suspicious cold solder connections, I just resolder
    the connections.

    I didn't go so far as to follow
    the traces to see where voltage went away. The owner is quite prosperous >>>and lives a distance away, and was saying "Dinner is ready!" She'll just >>>buy a new one. I figure I did my duty by eliminating obvious faults and >>>by pointing out that the blanket itself was still good, so just replace >>>the controller.

    We have a different view of the problem. I look my duty is to prevent
    yet another piece of almost working electronics from ending up in the >>recycling bin.

    "Recycling Isnt the Answer; Its the Last Resort" >><https://www.ifixit.com/Right-to-Repair/Recycling>

    Yes, I surmised that was the motivation. The terrible plague of electric >>>blanket deaths has finally been conquered! We are safer every year!

    Thank you for the cynical comment of the day.

    "Electric Blanket Fire Lawyer" >><https://www.killinofirm.com/practice-areas/electric-blanket-fire-lawyer> >>"Electric blankets cause about 5,000 house fires per year in the US, >>resulting in dozens of injuries and even death."

    But long before the days of transistors and "circuit boards" my folks
    had an electric blanket... used it all the rest of their lives. Is the
    modern one better?

    I've never slept under an electric blanket and I doubt I ever will. We
    raised sheep when I was growing up and we all had heavy woolen
    blankets. Our upstairs bedrooms were not heated and we sometimes
    needed two blankets. My family members and I still have some of them,
    although mine is packed away in a plastic case. It came from a
    Corriedale ewe that I'd bottled fed and won blue ribbons with at the
    4H fair.

    --
    C'est bon
    Soloman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Catrike Ryder@21:1/5 to AMuzi on Mon Jan 13 05:12:18 2025
    On Sun, 12 Jan 2025 19:28:34 -0600, AMuzi <am@yellowjersey.org> wrote:

    On 1/12/2025 7:19 PM, John B. wrote:
    On Sun, 12 Jan 2025 11:08:45 -0600, AMuzi <am@yellowjersey.org> wrote:

    On 1/11/2025 6:46 PM, Frank Krygowski wrote:
    On 1/11/2025 4:11 PM, Wolfgang Strobl wrote:
    Am Fri, 10 Jan 2025 20:01:26 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    On 1/10/2025 5:18 PM, Wolfgang Strobl wrote:
    This makes me reflect on the criticism of electronics in
    bicycles. ... But what about measuring the
    amount of power applied to the pedals, what about
    telling the cyclist
    who balanced he splits the power between left and right, by
    instrumenting the pedal or the bottom bracket?

    Yes, that can be done with enough sensors and
    electronics. But it seems
    like useless information to me. As with much computer
    software, it seems
    like "feature bloat." Why would anybody but a racer care?

    Is there any reason to believe that old people like me
    don't have a need
    to care about training intensity, as a matter of
    principle? How comes?

    When using my biycles as a middle aged, healthy adult I
    didn't care
    about racing, didn't train and didn't try to find company
    in cycling
    clubs, either. I didn't visit a gym until much later.
    Cycling was simply
    a way to get to work quickly. During vacations it was a more
    entertaining mode of transportation than the car. Fitness,
    strength and
    cardio fitness came as a side effect, slowly and almost
    unnoticed.

    This has changed. My cardio fitness is still better than
    what I know
    about a lot of people half my age. Nevertheless, it
    decreased over the
    years and it takes more and more effort to keep what you
    have not yet
    lost. As it is expected. The ability to measure power is
    helpful in
    finding the point at which intensity is sufficient and
    overload has not
    yet begun. This is even more true in combination with a
    heart rate
    monitor.

    So far, this is only about aging. But there is more.

    Accidents that led to damage to bones, joints and tendons
    in the past
    have consequences to be considered, too. There is a
    tendency to
    compensate weaknesses by bad postures, whithout noticing.
    This is both
    caused by damages and causing damages. This is to be
    avoided. Measuring
    how power is applied by the feet helps detecting
    differences early,
    avoiding damage.

    Avoiding damage is far from useless.

    But individual power outputs from left and right legs? I
    can't imagine much value from that measurement. In fact, I
    don't believe power measurement is necessary for fitness. I
    judge by feel. It has the advantage of naturally
    compensating for days when I feel stronger or weaker. And
    it's organic! ;-)

    My experiences with mechanical devices are at least as
    mixed as those
    with electronic devices. Some very complicated electronic
    devices that
    I bought decades ago still work. The same applies to some
    that I built
    myself a long time ago. With devices that contained both
    electronic and
    mechanical parts, the fault was more often on the
    mechanical side.

    Anecdotical, I know.

    To me, a big advantage is the ability to _look_ at a
    mechanical device and _see_ what's wrong. That, and the fact
    I can often affect a repair. I hate the Kleenex ethic -
    "It's no good any more, just throw it away."

    Two days ago, my kid asked me to figure out why an
    electric blanket
    wasn't working. The controller refused to turn on. I
    opened it and
    confirmed that it was getting supply voltage. Beyond
    that, the pile of
    dozens of surface mount electronic components was
    incomprehensible to
    me. I suspect Jeff might have been able to diagnose it,
    but not me.
    Here's a photo:
    https://www.flickr.com/photos/16972296@N08/54259119364/
    in/dateposted-public/
    I know electric blankets are old technology. I know they
    functioned well
    for decades with maybe a rheostat and perhaps a couple of
    other
    components. Why add unrepairable complexity?

    Don't know. We use devices like the blue one in the
    following picture.
    <https://upload.wikimedia.org/wikipedia/commons/4/4b/
    W%C3%A4rmflasche1.jpg>
    and better isolating blankets. No electricity necessary. :-)

    :-) But you imply that _I'm_ the one insufficiently modern?

    A few lights in our house are switched by set of 2 x 3
    inexpensive
    wireless sockets including two remote controls, that I
    bought eleven
    years ago. I've still to replace the batteries. Two of the
    sockets are
    still spares, I have a replacement cell for the remote
    controls stored
    which might live even longer. Standard type, used in
    garage openers and
    burglar alarms, too. Selecting a channel and paring one of
    the four
    buttons of a remote control with one or more of the
    sockets is as easy
    as pie, using a line of dip switches inside those
    devices. Quite
    similar to pairing switches and derailleur on our bicyles.

    While I avoid having essential functionality in my house
    depend on
    wireless connections, I enjoy having the option, for
    certain use cases,
    though.

    Yuck.

    I bought a new pedestal fan for our back patio last year.
    I've found a fan to be the best defense against mosquitos.
    The previous one still runs, but exposure to sunlight
    destroyed its plastic finger protection screen.

    Anyway, the new one comes with a remote control, roughly the
    size of a snack cracker. Who needs a remote to adjust a fan?
    I rejected ceiling fans with remotes when I was shopping,
    too. It's something to lose, something whose battery will
    die, and something whose electronics will go bad and be
    unrepairable.
    ...

    When I see a 50 year old, perfect looking bicycle, I think:
    That one must have been standing around inside most of the
    time, perhaps
    for a reason.

    My favorite bike is a young 39 years. It looks perfect, in
    part because I had it (and my wife's matching one) stripped
    and powder coated a couple years ago. Too many excellent
    memories to let it go.



    Not only the electronics infestation of consumer products.

    I brought food to make girlfriend's birthday dinner last
    evening and I cook in her kitchen regularly but sometimes
    the modern kitchen stove is a real impediment.

    [break for electronics rant: Her new kitchen stove, as her
    furnace, both run from the large LP tank in the yard. Both
    'improved' designs cannot run without electronic
    start/valves etc so when the power is out there is no heat
    whatsoever and power outages in rural USA are periodic.]

    The new kitchen stove has electronic controls for the
    burners and oven (oven controlled by a touch pad not, a
    physical switch). The burner controls have 4 positions
    only, viz., Off, Low, High, Start. That's really difficult
    for some cooking projects. Resolved by moving the two pans
    off and on the flame, which is Just One More Thing when
    timing two dishes at once.

    Only a designer who has never actually cooked would think of
    that. Works fine for brewing coffee or boiling pasta but
    limited for many projects.

    Gee,,, my mother cooked on a "gas" stove with mechanical controls from
    the regulate at the tank to the knobs on the stove successfully and
    one grandmother cooked on a wood stove, in the winter, un till she
    died, successfully.

    Are these "modern contraptions" really necessary?

    Well, perhaps they are. Are there young woman, today, being taught by
    their mother how to cook on a wood stove? Or how to darn woolen socks,
    or even how to raise the kids?


    Much agreed.

    The kitchen stove I used for decades had simple gas valves
    and a box of wooden matches in a clay dish next to it. No
    complaints.

    I did live for a while on the boat without a microwave, but I'll never
    do that again.

    --
    C'est bon
    Soloman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Wolfgang Strobl@21:1/5 to All on Mon Jan 13 09:28:46 2025
    Am Sun, 12 Jan 2025 15:23:33 -0800 schrieb Jeff Liebermann
    <jeffl@cruzio.com>:

    On Fri, 10 Jan 2025 23:18:27 +0100, Wolfgang Strobl

    ...

    My first and only HP calculator is a HP 200LX, still working fine.

    I had an HP95, HP 100LX, and a HP200LX. I still have the HP100-LX,
    but the others were sold long ago. Trying to find PCMCIA cards with
    small amounts of memory, is what forced me to stop using them. They
    were really "cool" in their day.

    I
    lost one of the tiny case screws, decades ago, replacing the CR2032 now
    and then is still awkward. I have no real use for the device anymore,
    but I still like it enough for not giving it away. AFAIR, there's still
    a copy of the original DeSmet C compiler on the flash card in the PCMCIA >>slot. :-)

    Ok, you're addicted. I have a rather disorganized box of HP
    calculator parts. I'll look for a screw and mail it to you.
    Otherwise, there seem to be some available online: ><https://www.200lx.net/mtncekit.htm>

    I removed one screw from my HP 100LX to check the size. I would
    guess(tm) it's 2-56, but without a gauge or a 2-56 nut, I can't be
    sure. However, I can't find my thread gauges and my microscope
    threading graticule is in use. So this will need to wait. Please
    remind me if I forget.

    Thanks for the kind offer. But please don't remove a screw from one of
    your devices. As I wrote, I haven't used my 200LX anymore for years. I
    enjoy having it in my collection, I keep it working just out of
    curiousity and sometimes to show younger people a device that no longer
    exists in this form today. The forth screw isn't necessary for that.


    Getting back to cycling ... :-)

    Sorry. I need to get back to house cleaning and firewood hauling.
    I'll read through the rest of your posting and write something,
    hopefully in a few days.

    I ment "getting back to talking about cycling". :-) But don't hesitate
    to comment. I'm all ears.

    Unfortunately, cycling outside the house will have to wait for some more months, so the indoor training must do :-/ In its own way, my own Tacx
    trainer is a construct of mechanics, electronics, computers and software
    too, that didn't exist a few decades ago and that some older people
    can't get used to.

    On the other hand, in the past few months I've met people who are even
    older than me and have no problems using larger versions and variants of
    what I have and use at home.

    <https://milongroup.com/en/circuit-training>


    --
    Thank you for observing all safety precautions

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From zen cycle@21:1/5 to Jeff Liebermann on Mon Jan 13 08:57:18 2025
    On 1/13/2025 2:01 AM, Jeff Liebermann wrote:
    On Mon, 13 Jan 2025 13:20:04 +0700, John B. <slocombjb@gmail.com>
    wrote:

    But long before the days of transistors and "circuit boards" my folks
    had an electric blanket... used it all the rest of their lives. Is the
    modern one better?

    Not really better, but certainly safer. Electric heaters are all 100% efficient. Every watt you shove into the heating wires is turned into
    heat. There's no way to improve on 100% efficiency.

    Nothing is 100% efficient


    The problem with electric blankets is that they're rather fragile.
    Dogs, cats and rats chew on them causing exposed wires. Using them
    folded can break wires at the bend or cause overheating. There are
    probably other ways to damage the wires. Once exposed and/or broken,
    the safety question becomes what will the circuitry to prevent sparks, localized overheating, starting a fire, overload, etc. Dealing with
    all those potential problems doesn't create a better heating
    experience, but does make the blanket safer. Basically, you don't
    know that you have a problem until after the house burns down.

    Electric heated blanket recalls: <https://www.cpsc.gov/search?search_api_fulltext=electric+blankets>
    If you check the "Include archived items" box, you'll see all the old recalls.



    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From zen cycle@21:1/5 to Frank Krygowski on Mon Jan 13 08:54:31 2025
    On 1/12/2025 2:25 PM, Frank Krygowski wrote:
    On 1/12/2025 7:49 AM, zen cycle wrote:
    On 1/11/2025 7:46 PM, Frank Krygowski wrote:

    But individual power outputs from left and right legs? I can't
    imagine much value from that measurement.

    You answered your own question. "You" can't imagine it's useful. There
    are a great many data geeks who find it quite useful.

    Maybe we can take a poll? How many here gather and use data on their individual legs' power output?



    I hardly consider this forum to be representative of the cycling
    community at alrge.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Frank Krygowski on Mon Jan 13 08:20:47 2025
    On 1/12/2025 8:05 PM, Frank Krygowski wrote:
    On 1/12/2025 3:33 PM, Wolfgang Strobl wrote:
    Am Sat, 11 Jan 2025 19:46:50 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    To me, a big advantage is the ability to _look_ at a
    mechanical device
    and _see_ what's wrong....

    That, and the fact I can often affect a repair.

    I prefer devices that don't need repair over their lifetime.

    The weakness I see with that is the assumption that
    "lifetime" is defined as "the amount of time it works." if
    something stops working, its lifetime is over! Throw it out!

    As I said, I hate the Kleenex ethic - "It's no good any
    more, just throw it away."

    Don't know.  We use devices like the blue one in the
    following picture.
    <https://upload.wikimedia.org/wikipedia/commons/4/4b/
    W%C3%A4rmflasche1.jpg>
    and better isolating blankets. No electricity
    necessary. :-)

    :-) But you imply that _I'm_ the one insufficiently modern?

    No. I doubt that "being modern" is a reasonable benchmark
    or measure.
    This cuts both ways. :-)

    A few lights in our house are switched by set of 2 x 3
    inexpensive
    wireless sockets including two remote controls, that I
    bought eleven
    years ago. I've still to replace the batteries. Two of
    the sockets are
    still spares, I have a replacement cell for the remote
    controls stored
    which might live even longer. Standard type, used in
    garage openers and
    burglar alarms, too. Selecting a channel and paring one
    of the four
    buttons of a remote control with one or more of the
    sockets is as easy
    as pie, using a line of dip switches inside those
    devices.  Quite
    similar to pairing switches and derailleur on our bicyles.

    While I avoid having essential functionality in my house
    depend on
    wireless connections, I enjoy having the option, for
    certain use cases,
    though.

    Yuck.

    You don't have a single remote control in your house? Not
    even for the
    TV? That's rare.

    I've got more remotes than I want. The TV's power button and
    channel changing buttons are not even visible. They're
    hidden and practically un-labeled on the back edge of the
    device, so a remote is necessary to even turn it on. I pump
    the TV sound through our stereo amplifier, which has its own
    remote (whose volume control seems to have stopped working),
    the CD/DVD player has a separate remote, etc. etc. If we had
    a friend house sit for us, I'd have to write a manual on how
    to run the system.

    A couple years ago we were given a Christmas gift of battery
    powered LED "candles." The could change colors - by use of a
    remote! Why should a candle need a remote?

    Same for a ceiling fan. Ours change speed by use of a pull
    chain. I'll never lose the pull chain. I would certainly
    misplace a remote.

    I still have a box of old, but still working IR remote
    controls,
    from devices that broke many years ago.  I collected these
    to control
    gadgets like this one
    <https://www.mystrobl.de/Plone/basteleien/microcontroller/
    ws2812/DSC_3564-DSC_3566_fused.JPG>
    Can you guess what this blinkenlight does?

    Nope.

    Parts a an IR receiver, a PIC 12F1840 microcontroller, a
    stripe of eight
    RGB LED, and a remote control from a CD Player that broke
    long before
    2014, when I built that gadget.  The aforementioned $1.50
    controller
    (single quantity, digikey) does everything from IR
    decoding to
    controlling the LED stripe.

    Somewhat later, I built something larger using a different
    part (an
    ESP8266) for illuminating the house bar of one of our
    kids, using about
    one meter of densely placed RGB LED, controlled via WIFI,
    doing a whole
    series of different colorful light effects. Extendable
    with new effects
    by uploading short LUA snippets, of course. :-) I was told
    that it was
    used again at a New Year's celebration, so obviously it is
    still
    working.

    There's a part of me that wishes I had your skills and
    knowledge. I occasionally dream up little electronic
    projects that I lack the knowledge to design or build. I've
    thought about educating myself, but soon realized there are
    many other things I'd prefer to learn.

    My wife still uses an almost as old bicycle for everyday
    rides around
    the corner.  Didn't have to strip and power coat it,
    because it came
    that way, when she bought it.

    But like me, she is glad that I build two road bikes in
    2023, using
    wireless electronic shifting that you dislike so much.
    Without, she
    wouldn't have been able to do some of those very enjoyable
    tours
    throughout the region that we did in 2023 and in 2024.

    Can you explain? It's hard for me to visualize a tour that
    would _require_ electronic shifting. The vast majority of my
    touring and riding miles have been done without even index
    shifting.


    That's very personal; some riders prefer fixed (although not
    this morning. It's bitter cold with a wind here).

    For riders with prior wrist/finger injuries, advancing
    arthritis and/or general age related decrepitude, Ergo/STi
    are difficult if not painful. For most riders, they're
    perfectly adequate. As are downtube friction shifters for
    others.

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Rolf Mantel@21:1/5 to All on Mon Jan 13 15:42:23 2025
    Am 13.01.2025 um 08:01 schrieb Jeff Liebermann:
    On Mon, 13 Jan 2025 13:20:04 +0700, John B. <slocombjb@gmail.com>
    wrote:

    But long before the days of transistors and "circuit boards" my folks
    had an electric blanket... used it all the rest of their lives. Is the
    modern one better?

    Not really better, but certainly safer. Electric heaters are all 100% efficient. Every watt you shove into the heating wires is turned into
    heat. There's no way to improve on 100% efficiency.

    Wrong: There's no *simple* way to improve on 100% efficiency. As soon
    as you change your definition of "efficiency", there might be additional improvements.
    Examples include
    <https://en.wikipedia.org/wiki/Condensing_boiler>
    (you gain additional heat by emitting liquid rather then gaseous H2O)

    or
    <https://en.wikipedia.org/wiki/Heat_pump>
    (cooling the outside brings you additional warmth inside).

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Wolfgang Strobl@21:1/5 to All on Mon Jan 13 15:57:24 2025
    Am Sun, 12 Jan 2025 21:05:47 -0500 schrieb Frank Krygowski <frkrygow@sbcglobal.net>:

    On 1/12/2025 3:33 PM, Wolfgang Strobl wrote:
    Am Sat, 11 Jan 2025 19:46:50 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    To me, a big advantage is the ability to _look_ at a mechanical device
    and _see_ what's wrong....

    That, and the fact I can often affect a repair.

    I prefer devices that don't need repair over their lifetime.

    The weakness I see with that is the assumption that "lifetime" is
    defined as "the amount of time it works." if something stops working,
    its lifetime is over! Throw it out!

    That's far too simplistic.

    It depends. For my purposes, I indeed prefer bicycles that may need
    repairs and modifications over their lifetime, for various reasons. I
    change over my lifetime, so do my bicycles. But there are limits. Want
    it cheap, longlived, lightweight and functional? Choose any two.



    As I said, I hate the Kleenex ethic - "It's no good any more, just throw
    it away."

    A strawman isn't getting any more pretty, over time. You won't find many complex products, machines, vehicles or components with an unlimited
    lifetime. Product lifetime has to be planned. There is innovation,
    innovation means change. There are technical limits. So far, I haven't
    heard about bicycle tires that tolerate heavy use over a lifetime of 40
    years, as you ask for. To be precise, I don't know of any that I would
    like to use or that I would risk using.

    But I still prefer devices that don't need repair over their lifetime,
    because all relevant components have an expected lifetime large enough
    for the expected combined lifetime to exceed a large planned lifetime.

    Sometimes that goal is hard or expensive to archive. Take the example
    above, tires. If some necessary component has a limited lifetime,
    standardize its interface, make it replaceable and guarantee that
    replacements will remain available for the lifetime of the product.

    Nothing stops people from repairing defective components, within limits.
    But that isn't the point. I guess even you don't use retreaded tires on
    your car, anymore. You probably won't repair a worn chain or chainring.

    Anyway, I see no reason why the wireless shifting of our bikes shouldn't outlive a similar purely mechanical one, simply because it produces less mechanical stress. Of course, SRAM or Shimano might artificially limit
    the lifetime of those products. Software and forced updates make that
    easier than it should be. But nothing prevents them from doing that on
    the mechanical side, either. There is no reason to blame a technology
    for that, just because you don't like it.



    Don't know. We use devices like the blue one in the following picture. >>>> <https://upload.wikimedia.org/wikipedia/commons/4/4b/W%C3%A4rmflasche1.jpg>
    and better isolating blankets. No electricity necessary. :-)

    :-) But you imply that _I'm_ the one insufficiently modern?

    No. I doubt that "being modern" is a reasonable benchmark or measure.
    This cuts both ways. :-)

    A few lights in our house are switched by set of 2 x 3 inexpensive
    wireless sockets including two remote controls, that I bought eleven
    years ago. I've still to replace the batteries. Two of the sockets are >>>> still spares, I have a replacement cell for the remote controls stored >>>> which might live even longer. Standard type, used in garage openers and >>>> burglar alarms, too. Selecting a channel and paring one of the four
    buttons of a remote control with one or more of the sockets is as easy >>>> as pie, using a line of dip switches inside those devices. Quite
    similar to pairing switches and derailleur on our bicyles.

    While I avoid having essential functionality in my house depend on
    wireless connections, I enjoy having the option, for certain use cases, >>>> though.

    Yuck.

    I really enjoy that I don't have to walk down two stairs just for the
    light I forgot to switch of in the living room. There have been other occasions when I was very happy to be able use a lamp whose switch I
    could not reach.



    You don't have a single remote control in your house? Not even for the
    TV? That's rare.

    I've got more remotes than I want.

    We aren't using those that came with our TV and those from our stereo
    system and amplifier, either. Our TV actually is a large monitor
    connected to a PC serving both as a DVB-T receiver and for internet
    streaming - viewing public television programs, that is. The monitor has
    an IR remote, but the remote control in use for the TV is just a
    wireless keyboard with a touchpad from Logitech. In the living room, to
    play a CD, listen to the radio, or play music from my cell phone,
    laptop, etc., I simply walk to the closet where the amplifier is
    located. If I'm lazy, I just use the phone and the BT connection to the amplifier.


    The TV's power button and channel
    changing buttons are not even visible. They're hidden and practically >un-labeled on the back edge of the device, so a remote is necessary to
    even turn it on.

    There was a similar problem with our TV, too many separate components. I
    solved that by using a power strip combined with a separate central
    switch at an easy to reach location. Powering on/off needs two actions: central switch plus a button on the PC, powering off is done via
    keyboard and central switch. That way, all that stuff doesn't consume
    standby power, when not in use.


    I pump the TV sound through our stereo amplifier, which
    has its own remote (whose volume control seems to have stopped working),
    the CD/DVD player has a separate remote, etc. etc. If we had a friend
    house sit for us, I'd have to write a manual on how to run the system.

    This can actually be automated quite easily for devices with IR remote controls. However, it does require a little programming and soldering
    work.


    A couple years ago we were given a Christmas gift of battery powered LED >"candles." The could change colors - by use of a remote! Why should a
    candle need a remote?

    Because you don't want to need a ladder to reach the point where it
    stands, just to operate the on/off switch?


    Same for a ceiling fan. Ours change speed by use of a pull chain. I'll
    never lose the pull chain. I would certainly misplace a remote.

    I've seen many such fans, radiant heaters and the like, where the pull
    chain or drawstring had been lost or damaged. But I have rarely
    misplacted an IR remote, simply because there is no point in moving it
    out of the room where the controlled device is located.


    I still have a box of old, but still working IR remote controls,
    from devices that broke many years ago. I collected these to control
    gadgets like this one
    <https://www.mystrobl.de/Plone/basteleien/microcontroller/ws2812/DSC_3564-DSC_3566_fused.JPG>
    Can you guess what this blinkenlight does?

    Nope.

    It plays Mastermind faster and better than I do.


    Parts a an IR receiver, a PIC 12F1840 microcontroller, a stripe of eight
    RGB LED, and a remote control from a CD Player that broke long before
    2014, when I built that gadget. The aforementioned $1.50 controller
    (single quantity, digikey) does everything from IR decoding to
    controlling the LED stripe.

    Somewhat later, I built something larger using a different part (an
    ESP8266) for illuminating the house bar of one of our kids, using about
    one meter of densely placed RGB LED, controlled via WIFI, doing a whole
    series of different colorful light effects. Extendable with new effects
    by uploading short LUA snippets, of course. :-) I was told that it was
    used again at a New Year's celebration, so obviously it is still
    working.

    There's a part of me that wishes I had your skills and knowledge. I >occasionally dream up little electronic projects that I lack the
    knowledge to design or build. I've thought about educating myself, but
    soon realized there are many other things I'd prefer to learn.

    I got the basic skills and knowledge while we programmed mainframes
    using IBM ASM F assembler, in my job. Doing it minimalistically on the
    bare metal as I like it is even easier, nowadays, with 8-bit
    microcontrollers - if you have grown up with that stuff. But there are alternatives, Arduino comes to mind. I don't like it, but understand how
    and why it is popular.



    My wife still uses an almost as old bicycle for everyday rides around
    the corner. Didn't have to strip and power coat it, because it came
    that way, when she bought it.

    But like me, she is glad that I build two road bikes in 2023, using
    wireless electronic shifting that you dislike so much. Without, she
    wouldn't have been able to do some of those very enjoyable tours
    throughout the region that we did in 2023 and in 2024.

    Can you explain? It's hard for me to visualize a tour that would
    _require_ electronic shifting.

    Southwest of our home begins a low mountain range called Eifel. After retirement, I started to ride my bike into that region, as far as I
    could and enjoyed that very much. But I have limits and so has my wife.
    I could motivate her to accompany me on some of those tours, but
    couldn't motivate her to try some of the steeper or longer ones. This
    was one of the reasons to build new bikes in 2023.

    The following route (shown by the web version of brouter, a routing tool
    used for example in the OsmAND app for Android phones) is a very scenic,
    but steep segment/shortcut, which is part of a handfull of roundtrips
    that I have done and enjoyed, with my old bike. Unfortunately, it was
    just a bit too hard for her with her old bike.

    <https://brouter.de/brouter-web/#map=14/50.5165/6.9122/standard&lonlats=6.910437,50.510777;6.91813,50.518292&profile=fastbike>

    Doing a 12% hill at our age with insufficient gears and a prissy
    gearshift is dangerous, especially if you don't have prior experience.
    Even worse, if you have got a a bad experience by trying and failing, as
    she had.

    So designing the new bike wasn't only about low enough gears. Simple,
    fast and reliable shifting was at least as important. An electronic SRAM
    eTAP AXS mullet with a small chainring delivered exactly that. 1 x 12,
    10-52 on the back, initially 40 in front, now 32 teeth. Operated by just pushing one of two switches.

    I could do without, but frankly, I enjoy it too!

    The following collage shows her riding up to and through Winnen with the
    new bike. A few days later she asked me: what about trying it again?

    <https://www.mystrobl.de/ws/pic/fahrrad/20230906/collage2.jpg>

    The vast majority of my touring and
    riding miles have been done without even index shifting.

    We too didn't start cycling just recently, either. :-) But unlike me,
    she didn't get fat skin and thick calves by commuting in heavy traffic
    and our long rides on vacation were rather easy. Not to mention, when we
    did our first tours with Peugeot bicycles bought in 1978, we were young
    and we mostly rode on the flat. A piece of cake, even with only 2 x 5
    and without indexing.


    --
    Thank you for observing all safety precautions

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Wolfgang Strobl@21:1/5 to All on Mon Jan 13 16:40:11 2025
    Am Mon, 13 Jan 2025 08:54:31 -0500 schrieb zen cycle <funkmasterxx@hotmail.com>:

    On 1/12/2025 2:25 PM, Frank Krygowski wrote:
    On 1/12/2025 7:49 AM, zen cycle wrote:
    On 1/11/2025 7:46 PM, Frank Krygowski wrote:

    But individual power outputs from left and right legs? I can't
    imagine much value from that measurement.

    You answered your own question. "You" can't imagine it's useful. There
    are a great many data geeks who find it quite useful.

    Maybe we can take a poll? How many here gather and use data on their
    individual legs' power output?

    Thinking about the question, I noticed that it is somewhat loaded.
    Depending on how the power output of the rider is measured, you get
    separate values for both feet anyway. For example, when power is
    measured by the pedals and later combined in the device collecting the
    data. Even more, raw data usually is a series of many values over one revolution. So all that data is available anyway, there is no decision
    from the side of the rider to gather that data. Combining and
    visualizing that data is just a little bit of software, so why not
    giving the user the ability to view that? Some users indeed can make
    good use of that data, sometimes.

    At least on a Garmin Edge, nobody is forced to add a widget that
    displays that data on a visible page of a profile. If all that somebody
    wants is average Watts over 30 seconds, only adding a widget that
    displays just that is sufficient. AFAIK, there is no widget that
    displays values for left and right, during a ride. At least I haven't
    seen one in the standard set. Haven't checked whether it is in the data
    written to the log (i.a. an activity).




    I hardly consider this forum to be representative of the cycling
    community at alrge.

    Right. I just did a poll in my family, counting four people having a
    powermeter on one of their bikes, three of them a powermeter that
    measures both feet separately. But how do I count the two indoor
    trainers, that can do that? Is the ability to detect an imbalance on at
    least one device sufficient for a [x] check on the questionnaire? ;-)

    While I'm at it, does anybody know how the Tacx Neo 2 T does it? I have
    my suspicions, but haven't read anything about it yet.


    --
    Thank you for observing all safety precautions

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Catrike Ryder@21:1/5 to AMuzi on Mon Jan 13 11:05:44 2025
    On Mon, 13 Jan 2025 08:20:47 -0600, AMuzi <am@yellowjersey.org> wrote:

    On 1/12/2025 8:05 PM, Frank Krygowski wrote:
    On 1/12/2025 3:33 PM, Wolfgang Strobl wrote:
    Am Sat, 11 Jan 2025 19:46:50 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    To me, a big advantage is the ability to _look_ at a
    mechanical device
    and _see_ what's wrong....

    That, and the fact I can often affect a repair.

    I prefer devices that don't need repair over their lifetime.

    The weakness I see with that is the assumption that
    "lifetime" is defined as "the amount of time it works." if
    something stops working, its lifetime is over! Throw it out!

    As I said, I hate the Kleenex ethic - "It's no good any
    more, just throw it away."

    Don't know. We use devices like the blue one in the
    following picture.
    <https://upload.wikimedia.org/wikipedia/commons/4/4b/
    W%C3%A4rmflasche1.jpg>
    and better isolating blankets. No electricity
    necessary. :-)

    :-) But you imply that _I'm_ the one insufficiently modern?

    No. I doubt that "being modern" is a reasonable benchmark
    or measure.
    This cuts both ways. :-)

    A few lights in our house are switched by set of 2 x 3
    inexpensive
    wireless sockets including two remote controls, that I
    bought eleven
    years ago. I've still to replace the batteries. Two of
    the sockets are
    still spares, I have a replacement cell for the remote
    controls stored
    which might live even longer. Standard type, used in
    garage openers and
    burglar alarms, too. Selecting a channel and paring one
    of the four
    buttons of a remote control with one or more of the
    sockets is as easy
    as pie, using a line of dip switches inside those
    devices. Quite
    similar to pairing switches and derailleur on our bicyles.

    While I avoid having essential functionality in my house
    depend on
    wireless connections, I enjoy having the option, for
    certain use cases,
    though.

    Yuck.

    You don't have a single remote control in your house? Not
    even for the
    TV? That's rare.

    I've got more remotes than I want. The TV's power button and
    channel changing buttons are not even visible. They're
    hidden and practically un-labeled on the back edge of the
    device, so a remote is necessary to even turn it on. I pump
    the TV sound through our stereo amplifier, which has its own
    remote (whose volume control seems to have stopped working),
    the CD/DVD player has a separate remote, etc. etc. If we had
    a friend house sit for us, I'd have to write a manual on how
    to run the system.

    A couple years ago we were given a Christmas gift of battery
    powered LED "candles." The could change colors - by use of a
    remote! Why should a candle need a remote?

    Same for a ceiling fan. Ours change speed by use of a pull
    chain. I'll never lose the pull chain. I would certainly
    misplace a remote.

    I still have a box of old, but still working IR remote
    controls,
    from devices that broke many years ago. I collected these
    to control
    gadgets like this one
    <https://www.mystrobl.de/Plone/basteleien/microcontroller/
    ws2812/DSC_3564-DSC_3566_fused.JPG>
    Can you guess what this blinkenlight does?

    Nope.

    Parts a an IR receiver, a PIC 12F1840 microcontroller, a
    stripe of eight
    RGB LED, and a remote control from a CD Player that broke
    long before
    2014, when I built that gadget. The aforementioned $1.50
    controller
    (single quantity, digikey) does everything from IR
    decoding to
    controlling the LED stripe.

    Somewhat later, I built something larger using a different
    part (an
    ESP8266) for illuminating the house bar of one of our
    kids, using about
    one meter of densely placed RGB LED, controlled via WIFI,
    doing a whole
    series of different colorful light effects. Extendable
    with new effects
    by uploading short LUA snippets, of course. :-) I was told
    that it was
    used again at a New Year's celebration, so obviously it is
    still
    working.

    There's a part of me that wishes I had your skills and
    knowledge. I occasionally dream up little electronic
    projects that I lack the knowledge to design or build. I've
    thought about educating myself, but soon realized there are
    many other things I'd prefer to learn.

    My wife still uses an almost as old bicycle for everyday
    rides around
    the corner. Didn't have to strip and power coat it,
    because it came
    that way, when she bought it.

    But like me, she is glad that I build two road bikes in
    2023, using
    wireless electronic shifting that you dislike so much.
    Without, she
    wouldn't have been able to do some of those very enjoyable
    tours
    throughout the region that we did in 2023 and in 2024.

    Can you explain? It's hard for me to visualize a tour that
    would _require_ electronic shifting. The vast majority of my
    touring and riding miles have been done without even index
    shifting.


    That's very personal; some riders prefer fixed (although not
    this morning. It's bitter cold with a wind here).

    For riders with prior wrist/finger injuries, advancing
    arthritis and/or general age related decrepitude, Ergo/STi
    are difficult if not painful. For most riders, they're
    perfectly adequate. As are downtube friction shifters for
    others.

    My first external geared bicycles had friction shifters that I never
    mastered to where I didn't have to check by looking or listening. I
    don't remember which of my subsequent bicycle upgrades was the first
    to use index shifting, but I know I know the Schwinn (I think it was a
    Varsity) had it because I remember working on dialing it in. It wasn't
    until the summer of '23 that I got the trigger shifters. It took a few
    tries before I figured out to use them on my handlebar configuration,
    they are a definite improvement in that am taking more care at
    intersections by slowing and/or stopping.

    --
    C'est bon
    Soloman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Frank Krygowski on Mon Jan 13 10:36:32 2025
    On 1/13/2025 10:06 AM, Frank Krygowski wrote:
    On 1/12/2025 10:33 PM, Jeff Liebermann wrote:
    On Sun, 12 Jan 2025 21:11:37 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/12/2025 5:13 PM, Jeff Liebermann wrote:
    On Fri, 10 Jan 2025 20:01:26 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    Two days ago, my kid asked me to figure out why an
    electric blanket
    wasn't working. The controller refused to turn on. I
    opened it and
    confirmed that it was getting supply voltage. Beyond
    that, the pile of
    dozens of surface mount electronic components was
    incomprehensible to
    me. I suspect Jeff might have been able to diagnose it,
    but not me.
    Here's a photo:
    https://www.flickr.com/photos/16972296@N08/54259119364/
    in/dateposted-public/

    I don't think I can do much with just a photograph.  If
    you have the
    maker, model and FCC ID number, I can do better.  Also,
    some
    indication of what it does when the power is applied,
    such as do the
    LED's light.  If the don't light, try again in the dark
    and see if
    they partially light.

    There was no action at all, no lights, nothing.

    That makes troubleshooting easy.  There has to be a fuse
    or thermal
    circuit breaker in there somewhere.  I can't see one on
    the PCB
    (printed circuit board).  It's a single sided PCB so it's
    unlikely
    that it's hiding on the back of the PCB.  It's either
    inside the wall
    plug, inside some kind of on/off switch box that's in
    series with the
    power cord, or attached to the blanket somewhere.

    I did measure 120V on the board where the input wires were
    soldered in place.

    Sorry about the quality of the photo. I took the photo only
    to show the owner why I wasn't digging deeper into diagnosis.

    If you want me to continue, please provide the maker,
    model number,
    and FCC ID if available.

    I think you shouldn't bother, but I'll be visiting again
    this week. If the old unit is still around, maybe I'll bring
    it home for more detective work.


    One ever popular technique to resolve intractable electronics:

    https://www.yellowjersey.org/photosfromthepast/FIXCOMPU.JPG


    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Wolfgang Strobl@21:1/5 to All on Mon Jan 13 18:06:37 2025
    Am Mon, 13 Jan 2025 11:03:03 -0500 schrieb Frank Krygowski <frkrygow@sbcglobal.net>:

    On 1/13/2025 8:57 AM, zen cycle wrote:
    On 1/13/2025 2:01 AM, Jeff Liebermann wrote:
    On Mon, 13 Jan 2025 13:20:04 +0700, John B. <slocombjb@gmail.com>
    wrote:

    But long before the days of transistors and "circuit boards" my folks
    had an electric blanket... used it all the rest of their lives. Is the >>>> modern one better?

    Not really better, but certainly safer.  Electric heaters are all 100%
    efficient.  Every watt you shove into the heating wires is turned into
    heat.  There's no way to improve on 100% efficiency.

    Nothing is 100% efficient

    It's certainly true that 100% of the electricity consumed by an electric >blanket becomes heat.

    But that 100% isn't efficiency, it's just a raw conversion rate for
    Electricity to heat in an isolated blanket.

    You, as a person, don't need or consume heat. Somewhat simplified, you
    need a certain range of temperatures. In the ideal case, you don't need
    any additional energy, because your body already produces heat. A little
    bit of isolation, perhaps provided by that very blanket, might be
    sufficient. Actually, getting rid of that heat can become difficult,
    and this is getting worse, over time.


    I've always had a mindset for minimizing waste. For most of my life I
    was diligent about shutting off the light switch when leaving a room.
    But now I often don't bother, especially in winter. LEDs draw so much
    less electricity anyway, and in the winter the "waste" just amounts to >electric heat.

    On the other hand, producing LED generates waste too and many LED lamps
    aren't as long lived as the packing says. Especially LED lights
    compatible to former light bulb are prone to early failures, because te
    base doesn't provide enough space for reliable electronics. So
    overheating the electronics and undersized capacitors might kill some
    older and/or cheap LED bulbs almost as fast as the incandescent bulbs of
    the past that got replaced. It has gotten better over time though, so
    this is no argument for bulbs that only convert 5-10% of electric energy
    to light. But it is an argument not to waste light and so cancel out
    the benefit.


    <https://www.sciencedirect.com/science/article/abs/pii/S0301421514002638>

    "A brighter future? Quantifying the rebound effect in energy efficient lighting"

    <https://www.arquiled.com/en/avoiding-the-rebound-effect-when-transitioning-to-led/>

    "In Portugal, the average amount of light has risen 120% in the last
    five years. It is urgent to prevent the energy savings associated with
    the transition to LED from being offset by unnecessary lighting"

    --
    Wir danken für die Beachtung aller Sicherheitsbestimmungen

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Catrike Ryder@21:1/5 to frkrygow@sbcglobal.net on Mon Jan 13 12:46:44 2025
    On Mon, 13 Jan 2025 12:38:29 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/13/2025 8:54 AM, zen cycle wrote:
    On 1/12/2025 2:25 PM, Frank Krygowski wrote:
    On 1/12/2025 7:49 AM, zen cycle wrote:
    On 1/11/2025 7:46 PM, Frank Krygowski wrote:

    But individual power outputs from left and right legs? I can't
    imagine much value from that measurement.

    You answered your own question. "You" can't imagine it's useful.
    There are a great many data geeks who find it quite useful.

    Maybe we can take a poll? How many here gather and use data on their
    individual legs' power output?

    I hardly consider this forum to be representative of the cycling
    community at alrge.

    I certainly agree with that! This is a group with no members who were
    not interested in the title "tech". So if these people are not
    measuring individual leg power, I doubt it's very common.

    Our bike club is mostly social, not very competitive. I do have friends
    who use "new" equipment (e.g. aero wheels, 13 cogs, road disc brakes)
    and a few who mention their KOM triumphs. Many of them use Garmin or
    similar devices. But IIRC none have ever discussed power meter readings. >Maybe some new, younger club members are doing that, but I don't ride
    with those folks.

    Then there's the vast majority of bike riders who will never join a
    club, never look at a bike magazine, never try to ride fast, who just
    ride for fun. No power meters there.

    But the bigger question is, "why do you care if others are using power
    meters?"

    --
    C'est bon
    Soloman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to funkmasterxx@hotmail.com on Mon Jan 13 11:28:12 2025
    On Mon, 13 Jan 2025 08:57:18 -0500, zen cycle
    <funkmasterxx@hotmail.com> wrote:

    On 1/13/2025 2:01 AM, Jeff Liebermann wrote:
    On Mon, 13 Jan 2025 13:20:04 +0700, John B. <slocombjb@gmail.com>
    wrote:

    But long before the days of transistors and "circuit boards" my folks
    had an electric blanket... used it all the rest of their lives. Is the
    modern one better?

    Not really better, but certainly safer. Electric heaters are all 100%
    efficient. Every watt you shove into the heating wires is turned into
    heat. There's no way to improve on 100% efficiency.

    Nothing is 100% efficient

    Well, the heater wires might radiate some EM radiation, which then is
    absorbed by nearby objects and is converted to IR which adds to the
    heat produced by the heater wires. There might be some radiation at
    other frequencies (RFI, EMI, microwaves, visible light, UV, etc) but
    most of the radiation is ends up somewhere in the IR bands.

    Incidentally, I'm usually amused at the advertising for electric
    heaters all claiming that their more "efficient" than the competition.
    Of course, no numbers and certainly no calculations are ever provided.

    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Catrike Ryder@21:1/5 to All on Mon Jan 13 14:33:13 2025
    On Mon, 13 Jan 2025 11:28:12 -0800, Jeff Liebermann <jeffl@cruzio.com>
    wrote:

    On Mon, 13 Jan 2025 08:57:18 -0500, zen cycle
    <funkmasterxx@hotmail.com> wrote:

    On 1/13/2025 2:01 AM, Jeff Liebermann wrote:
    On Mon, 13 Jan 2025 13:20:04 +0700, John B. <slocombjb@gmail.com>
    wrote:

    But long before the days of transistors and "circuit boards" my folks
    had an electric blanket... used it all the rest of their lives. Is the >>>> modern one better?

    Not really better, but certainly safer. Electric heaters are all 100%
    efficient. Every watt you shove into the heating wires is turned into
    heat. There's no way to improve on 100% efficiency.

    Nothing is 100% efficient

    Well, the heater wires might radiate some EM radiation, which then is >absorbed by nearby objects and is converted to IR which adds to the
    heat produced by the heater wires. There might be some radiation at
    other frequencies (RFI, EMI, microwaves, visible light, UV, etc) but
    most of the radiation is ends up somewhere in the IR bands.

    Incidentally, I'm usually amused at the advertising for electric
    heaters all claiming that their more "efficient" than the competition.
    Of course, no numbers and certainly no calculations are ever provided.

    They could always do a study where they ask people which heater made
    them feel warmer.

    --
    C'est bon
    Soloman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to frkrygow@sbcglobal.net on Mon Jan 13 11:48:30 2025
    On Mon, 13 Jan 2025 11:03:03 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/13/2025 8:57 AM, zen cycle wrote:
    On 1/13/2025 2:01 AM, Jeff Liebermann wrote:
    On Mon, 13 Jan 2025 13:20:04 +0700, John B. <slocombjb@gmail.com>
    wrote:

    But long before the days of transistors and "circuit boards" my folks
    had an electric blanket... used it all the rest of their lives. Is the >>>> modern one better?

    Not really better, but certainly safer. Electric heaters are all 100%
    efficient. Every watt you shove into the heating wires is turned into
    heat. There's no way to improve on 100% efficiency.

    Nothing is 100% efficient

    It's certainly true that 100% of the electricity consumed by an electric >blanket becomes heat.

    True (because that's what I said). However, there's some question
    about how well a heater will warm the user, room, environment,
    atmosphere, etc. This video compares the effectiveness of heating
    with a heater that uses convection with one that uses radiation. Both
    types are 100% efficient, but deliver and distribute the heat
    differently.

    "Is Electric Heating Really 100% Efficient?" <https://www.youtube.com/watch?v=B7X6wmBKG_4> (6:07)

    I've always had a mindset for minimizing waste. For most of my life I
    was diligent about shutting off the light switch when leaving a room.
    But now I often don't bother, especially in winter. LEDs draw so much
    less electricity anyway, and in the winter the "waste" just amounts to >electric heat.

    I'm not so diligent. I still use an old coil wire electric wall
    heater in the bathroom. It's literally falling apart and should have
    been replaced many years ago. Yet another project.

    I recently discovered that I had left a 12VDC power supply turned on
    but with nothing connected drawing 12 watts continuously. It's
    probably been like that for the last 10 years. I did the math and
    discovered that it was costing me $23/year to do nothing but add a
    little more heat to the room. My solutions was to turn it off.


    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Wolfgang Strobl@21:1/5 to All on Mon Jan 13 20:48:36 2025
    Am Mon, 13 Jan 2025 12:27:27 -0500 schrieb Frank Krygowski <frkrygow@sbcglobal.net>:

    On 1/13/2025 9:57 AM, Wolfgang Strobl wrote:
    Am Sun, 12 Jan 2025 21:05:47 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    On 1/12/2025 3:33 PM, Wolfgang Strobl wrote:
    Am Sat, 11 Jan 2025 19:46:50 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    To me, a big advantage is the ability to _look_ at a mechanical device >>>>> and _see_ what's wrong....

    That, and the fact I can often affect a repair.

    I prefer devices that don't need repair over their lifetime.

    The weakness I see with that is the assumption that "lifetime" is
    defined as "the amount of time it works." if something stops working,
    its lifetime is over! Throw it out!

    That's far too simplistic.

    It depends. For my purposes, I indeed prefer bicycles that may need
    repairs and modifications over their lifetime, for various reasons. I
    change over my lifetime, so do my bicycles. But there are limits. Want
    it cheap, longlived, lightweight and functional? Choose any two.



    As I said, I hate the Kleenex ethic - "It's no good any more, just throw >>> it away."

    A strawman isn't getting any more pretty, over time. You won't find many
    complex products, machines, vehicles or components with an unlimited
    lifetime. Product lifetime has to be planned. There is innovation,
    innovation means change. There are technical limits. So far, I haven't
    heard about bicycle tires that tolerate heavy use over a lifetime of 40
    years, as you ask for. To be precise, I don't know of any that I would
    like to use or that I would risk using.

    I think my Cannondale touring bike qualifies. Of course I've replaced >consumable items like tires, chains, cogs, brake shoes, handlebar tape
    and occasionally a chainring.

    That way, any bicycle qualifies. There is essentially no part of a
    bicycle that isn't "consumable". If you are lucky, all consumable items
    are consumed at the same time. So you can just buy another bike, call
    it the repaired one and throw out the old one. :-) Given that declaring something consumed is a rather arbitrary decision, you have quite a lot
    of slack with that.


    I've made some equipment substitutions
    (saddle, bar-end shifters, "aero" brake levers) but the original
    equipment is exceptionally durable.

    That is a tautology. Of course the remaining equipment is exeptionally
    durable, otherwise it would have been replaced earlier, for whatever
    reason.


    For one example: The square taper cranks that Tom mocks still work
    perfectly well. I had to replace the original sealed bottom bracket one
    time, but there was no confusion about compatibility (and my cranks did
    not fall off!). The Stronglite roller bearing headset has also lasted >decades, with one parts replacement. The SunTour rear derailleur is
    still perfect, although I did cheat a bit. When I powder coated our
    bikes, I traded my derailleur for my wife's, figuring hers had many
    fewer miles; but both still work just fine. Wheels are not original
    because I switched from 27" to 700C, but they're 20 years old.

    So why didn't you buy a 40 years old bicycle from somebody who doesn't
    need his bicycle anymore?

    I guess you don't drive a Ford Model T and you don't use an grammophone
    that needs a steel needle for playing shellac records.

    Personally, I am more concerned about how to use a bicycle rather than
    other modes of transport and optimising the bike for that purpose, and I
    am less concerndedabout whether the bike choosen it will last ten,
    twenty or thirty years.

    How long a bicycle lasts depends upon how much it is used and under what conditions. A bicycle that lasts more than thirty years is most likely a display piece. That some people like you have the time, space and energy
    to maintain a bicycle much longer than its useful life is under normal conditions doesn't prove the opposite. That is not an argument against repairing, but an argument against repairing, whatever the cost. I'm not talking about money only, here. I mostly miss a sense of proportion.



    Anyway, I see no reason why the wireless shifting of our bikes shouldn't
    outlive a similar purely mechanical one...

    I guess we'll see, eventually.

    If we don't try, we certainly won't see it. Try to see it the following
    way: _you_ don't have any reason to try a group with wireless shifting
    like the one I built our bikes with, I understand that. So just let
    people like us who experience, like and sometimes need the benefits pay
    the money, try this innovation, and serve as guinea pigs.


    There was a similar problem with our TV, too many separate components. I
    solved that by using a power strip combined with a separate central
    switch at an easy to reach location. Powering on/off needs two actions:
    central switch plus a button on the PC, powering off is done via
    keyboard and central switch. That way, all that stuff doesn't consume
    standby power, when not in use.


    I pump the TV sound through our stereo amplifier, which
    has its own remote (whose volume control seems to have stopped working), >>> the CD/DVD player has a separate remote, etc. etc. If we had a friend
    house sit for us, I'd have to write a manual on how to run the system.

    This can actually be automated quite easily for devices with IR remote
    controls. However, it does require a little programming and soldering
    work.

    About that: A few years ago I got annoyed at the number of remotes. I'd
    read a good review about a programmable universal remote, and bought it.
    I followed the tedious instructions to program it so I could hit one
    button for "Watch TV", another button for "Play CD", another button for >"Listen to radio" etc.

    It's less than ideal. Part of the problem, I think, is that some of the >devices use the same signal code as a toggle for "power-on" &
    "power-off", as opposed to a separate code for "On" and "Off." If a
    device is left in the wrong state, things don't work. There was also
    some dimly remembered problem where commands from the remote had to
    arrive at the TV at the proper instant - not too soon, not too late -
    and the program couldn't manage that, despite the nice lady at the 800
    help number trying over and over to cure. (I suppose I could dig back
    into the programming, but I'm not motivated.)

    Some years ago, I helped extending a library that implements both
    reading (decoding) and sending (generating) IR codes, the primary author
    was quite prolific in extending it to any protocol that he got
    specificattions and/or samples for. I only wrote a driver part for a not
    yet supported microcontroller, but that was good enough to understand
    some of its workings. Sadly, the project has mostly stalled after 2015.
    The code is still working and small microcontrollers and IR remotes
    don't change that much.

    From your description, I cannot deduce whether a single button press on
    your IR control serves as an "invert the boolean that denotes specific
    state" (power on/off, for example), or if it is something else. Some
    universal remotes are just simple and stupid recorders, recording and
    replaying a bitstream without decoding, perhaps after some signal
    cleanup. Others decode and work from tables.

    From memory, most IR remotes use the NEC protocol, after extracting an
    abstract code from the bistream, that code essentially is triple (device address, command number, modifier), device denoting a specific tv model,
    for example, command number some arbitrary numbering of the keys on the
    remote, modifier in this case just a single bit denoting "this is
    comming from a repeating, still pressed keys).

    Usually, the behaviour of a IR remote control is as simple as that.

    So much about the basics. Knowing neither your universal control, nor
    anything about the remote control in question, I can't even guess what
    is causing that problem. I could tell you what I would do to analyze
    and perhaps solve it, but that won't help you, because you don't have
    the necessary equipment (and knowledge). And even I have stored away
    most of the stuff I need for such work, in order to get space to build
    and maintain our bikes.




    I've seen many such fans, radiant heaters and the like, where the pull
    chain or drawstring had been lost or damaged. But I have rarely
    misplacted an IR remote, simply because there is no point in moving it
    out of the room where the controlled device is located.

    Your rooms must be much less messy than mine!

    Not really. It's just that the remotes sit on top of the devices that
    they control, when not in use. Most of the time, that is.


    ...when we
    did our first tours with Peugeot bicycles bought in 1978, we were young
    and we mostly rode on the flat. A piece of cake, even with only 2 x 5
    and without indexing.

    As I get older and older, I'm trying to get more comfortable with being
    an old guy who avoids mountains. Maybe some guy in a Frank Patterson
    drawing: >https://www.worthpoint.com/worthopedia/frank-patterson-cycling-artist-500010568

    Well yes, I get the feeling, believe me. I've ridden exactly 4.9 km
    outside, over the last five months. :-/ I'm trying hard not to become comfortable with that. :-) But that's a different story and not one for
    this forum. Just this much: I've never cycled as far and as high on a
    single day as I have done repeatedly since we retired. Becoming old is
    an obstacle, no question. But you can postpone the consequences of
    ageing, at least for a while.

    --
    Thank you for observing all safety precautions

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Wolfgang Strobl@21:1/5 to All on Mon Jan 13 21:36:18 2025
    Am Mon, 13 Jan 2025 12:58:35 -0500 schrieb Frank Krygowski <frkrygow@sbcglobal.net>:

    On 1/13/2025 12:06 PM, Wolfgang Strobl wrote:
    Am Mon, 13 Jan 2025 11:03:03 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    On 1/13/2025 8:57 AM, zen cycle wrote:
    On 1/13/2025 2:01 AM, Jeff Liebermann wrote:
    On Mon, 13 Jan 2025 13:20:04 +0700, John B. <slocombjb@gmail.com>
    wrote:

    But long before the days of transistors and "circuit boards" my folks >>>>>> had an electric blanket... used it all the rest of their lives. Is the >>>>>> modern one better?

    Not really better, but certainly safer.  Electric heaters are all 100% >>>>> efficient.  Every watt you shove into the heating wires is turned into >>>>> heat.  There's no way to improve on 100% efficiency.

    Nothing is 100% efficient

    It's certainly true that 100% of the electricity consumed by an electric >>> blanket becomes heat.

    But that 100% isn't efficiency, it's just a raw conversion rate for
    Electricity to heat in an isolated blanket.

    ISTM a good definition of efficiency is "Desired output divided by
    input." Since the desired output is heat, I think it's 100% efficient.

    "desired output" does some heavy lifting, here. What about the desired
    input?

    Textbook definitions from a specific context are often quite misleading,
    when used in a different context. Rolf Mantel just gave an example for
    that in <vm38of$1qe41$1@dont-email.me>, there are many more.



    You, as a person, don't need or consume heat. Somewhat simplified, you
    need a certain range of temperatures. In the ideal case, you don't need
    any additional energy, because your body already produces heat. A little
    bit of isolation, perhaps provided by that very blanket, might be
    sufficient.

    And, I suppose, we could do away with all home heating, and just wear
    very heavy clothing all winter. But I don't know of anyone doing that.

    Actually, we added some isolation to our house, last year. The reduced
    energy consumption was quite noticeable and better than we expected. No
    heavy clothing necessary. It's an old house, unfortunately some
    regulations prohibit doing more or would make it very expensive. Decades
    ago, a former colleague build a house according to current standards
    from that time, heated by a heat pump and geothermics. Compared to the
    quoted 100 % "efficiency" of your heated blanket, that heating is ~750% efficient. While riding over the land during in 2021ff, I saw not only
    many new collectors on the roofs, mostly photovoltaic now, people were isolating their freestanding houses to such an extent that you would
    hardly believe it if you hadn't seen it.




    <https://www.sciencedirect.com/science/article/abs/pii/S0301421514002638>

    "A brighter future? Quantifying the rebound effect in energy efficient
    lighting"

    <https://www.arquiled.com/en/avoiding-the-rebound-effect-when-transitioning-to-led/>

    "In Portugal, the average amount of light has risen 120% in the last
    five years. It is urgent to prevent the energy savings associated with
    the transition to LED from being offset by unnecessary lighting"

    I don't perceive that to be a serious problem, at least indoors.

    The difference between the expected and actual efficiency is independent
    of whether the lighting is installed indoors or outdoors.

    But there is indeed another and additional problem caused by too much
    light outside: Both astronomers and wildlife are not happy about the
    increasing light pollution. And, to be honest, I don't like the fact
    that you can hardly see the stars anywhere these days either.


    If the
    objective is to expend less energy, it's still been achieved.

    If you are allowed to move the goalpost, the goal is already guaranteed.


    I'm not a fan of overly bright headlamps nor outdoor light pollution,
    but those seem to be separate issues.

    That is true, but obviously no argument. Light pollution is just another detriment of wasting energy by unnecessary lighting.


    --
    Thank you for observing all safety precautions

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Wolfgang Strobl on Mon Jan 13 14:31:10 2025
    On 1/13/2025 1:48 PM, Wolfgang Strobl wrote:
    Am Mon, 13 Jan 2025 12:27:27 -0500 schrieb Frank Krygowski <frkrygow@sbcglobal.net>:

    On 1/13/2025 9:57 AM, Wolfgang Strobl wrote:
    Am Sun, 12 Jan 2025 21:05:47 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    On 1/12/2025 3:33 PM, Wolfgang Strobl wrote:
    Am Sat, 11 Jan 2025 19:46:50 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    To me, a big advantage is the ability to _look_ at a mechanical device >>>>>> and _see_ what's wrong....

    That, and the fact I can often affect a repair.

    I prefer devices that don't need repair over their lifetime.

    The weakness I see with that is the assumption that "lifetime" is
    defined as "the amount of time it works." if something stops working,
    its lifetime is over! Throw it out!

    That's far too simplistic.

    It depends. For my purposes, I indeed prefer bicycles that may need
    repairs and modifications over their lifetime, for various reasons. I
    change over my lifetime, so do my bicycles. But there are limits. Want
    it cheap, longlived, lightweight and functional? Choose any two.



    As I said, I hate the Kleenex ethic - "It's no good any more, just throw >>>> it away."

    A strawman isn't getting any more pretty, over time. You won't find many >>> complex products, machines, vehicles or components with an unlimited
    lifetime. Product lifetime has to be planned. There is innovation,
    innovation means change. There are technical limits. So far, I haven't
    heard about bicycle tires that tolerate heavy use over a lifetime of 40
    years, as you ask for. To be precise, I don't know of any that I would
    like to use or that I would risk using.

    I think my Cannondale touring bike qualifies. Of course I've replaced
    consumable items like tires, chains, cogs, brake shoes, handlebar tape
    and occasionally a chainring.

    That way, any bicycle qualifies. There is essentially no part of a
    bicycle that isn't "consumable". If you are lucky, all consumable items
    are consumed at the same time. So you can just buy another bike, call
    it the repaired one and throw out the old one. :-) Given that declaring something consumed is a rather arbitrary decision, you have quite a lot
    of slack with that.


    I've made some equipment substitutions
    (saddle, bar-end shifters, "aero" brake levers) but the original
    equipment is exceptionally durable.

    That is a tautology. Of course the remaining equipment is exeptionally durable, otherwise it would have been replaced earlier, for whatever
    reason.


    For one example: The square taper cranks that Tom mocks still work
    perfectly well. I had to replace the original sealed bottom bracket one
    time, but there was no confusion about compatibility (and my cranks did
    not fall off!). The Stronglite roller bearing headset has also lasted
    decades, with one parts replacement. The SunTour rear derailleur is
    still perfect, although I did cheat a bit. When I powder coated our
    bikes, I traded my derailleur for my wife's, figuring hers had many
    fewer miles; but both still work just fine. Wheels are not original
    because I switched from 27" to 700C, but they're 20 years old.

    So why didn't you buy a 40 years old bicycle from somebody who doesn't
    need his bicycle anymore?

    I guess you don't drive a Ford Model T and you don't use an grammophone
    that needs a steel needle for playing shellac records.

    Personally, I am more concerned about how to use a bicycle rather than
    other modes of transport and optimising the bike for that purpose, and I
    am less concerndedabout whether the bike choosen it will last ten,
    twenty or thirty years.

    How long a bicycle lasts depends upon how much it is used and under what conditions. A bicycle that lasts more than thirty years is most likely a display piece. That some people like you have the time, space and energy
    to maintain a bicycle much longer than its useful life is under normal conditions doesn't prove the opposite. That is not an argument against repairing, but an argument against repairing, whatever the cost. I'm not talking about money only, here. I mostly miss a sense of proportion.



    Anyway, I see no reason why the wireless shifting of our bikes shouldn't >>> outlive a similar purely mechanical one...

    I guess we'll see, eventually.

    If we don't try, we certainly won't see it. Try to see it the following
    way: _you_ don't have any reason to try a group with wireless shifting
    like the one I built our bikes with, I understand that. So just let
    people like us who experience, like and sometimes need the benefits pay
    the money, try this innovation, and serve as guinea pigs.


    There was a similar problem with our TV, too many separate components. I >>> solved that by using a power strip combined with a separate central
    switch at an easy to reach location. Powering on/off needs two actions: >>> central switch plus a button on the PC, powering off is done via
    keyboard and central switch. That way, all that stuff doesn't consume
    standby power, when not in use.


    I pump the TV sound through our stereo amplifier, which
    has its own remote (whose volume control seems to have stopped working), >>>> the CD/DVD player has a separate remote, etc. etc. If we had a friend
    house sit for us, I'd have to write a manual on how to run the system.

    This can actually be automated quite easily for devices with IR remote
    controls. However, it does require a little programming and soldering
    work.

    About that: A few years ago I got annoyed at the number of remotes. I'd
    read a good review about a programmable universal remote, and bought it.
    I followed the tedious instructions to program it so I could hit one
    button for "Watch TV", another button for "Play CD", another button for
    "Listen to radio" etc.

    It's less than ideal. Part of the problem, I think, is that some of the
    devices use the same signal code as a toggle for "power-on" &
    "power-off", as opposed to a separate code for "On" and "Off." If a
    device is left in the wrong state, things don't work. There was also
    some dimly remembered problem where commands from the remote had to
    arrive at the TV at the proper instant - not too soon, not too late -
    and the program couldn't manage that, despite the nice lady at the 800
    help number trying over and over to cure. (I suppose I could dig back
    into the programming, but I'm not motivated.)

    Some years ago, I helped extending a library that implements both
    reading (decoding) and sending (generating) IR codes, the primary author
    was quite prolific in extending it to any protocol that he got specificattions and/or samples for. I only wrote a driver part for a not
    yet supported microcontroller, but that was good enough to understand
    some of its workings. Sadly, the project has mostly stalled after 2015.
    The code is still working and small microcontrollers and IR remotes
    don't change that much.

    From your description, I cannot deduce whether a single button press on
    your IR control serves as an "invert the boolean that denotes specific state" (power on/off, for example), or if it is something else. Some universal remotes are just simple and stupid recorders, recording and replaying a bitstream without decoding, perhaps after some signal
    cleanup. Others decode and work from tables.

    From memory, most IR remotes use the NEC protocol, after extracting an abstract code from the bistream, that code essentially is triple (device address, command number, modifier), device denoting a specific tv model,
    for example, command number some arbitrary numbering of the keys on the remote, modifier in this case just a single bit denoting "this is
    comming from a repeating, still pressed keys).

    Usually, the behaviour of a IR remote control is as simple as that.

    So much about the basics. Knowing neither your universal control, nor anything about the remote control in question, I can't even guess what
    is causing that problem. I could tell you what I would do to analyze
    and perhaps solve it, but that won't help you, because you don't have
    the necessary equipment (and knowledge). And even I have stored away
    most of the stuff I need for such work, in order to get space to build
    and maintain our bikes.




    I've seen many such fans, radiant heaters and the like, where the pull
    chain or drawstring had been lost or damaged. But I have rarely
    misplacted an IR remote, simply because there is no point in moving it
    out of the room where the controlled device is located.

    Your rooms must be much less messy than mine!

    Not really. It's just that the remotes sit on top of the devices that
    they control, when not in use. Most of the time, that is.


    ...when we
    did our first tours with Peugeot bicycles bought in 1978, we were young
    and we mostly rode on the flat. A piece of cake, even with only 2 x 5
    and without indexing.

    As I get older and older, I'm trying to get more comfortable with being
    an old guy who avoids mountains. Maybe some guy in a Frank Patterson
    drawing:
    https://www.worthpoint.com/worthopedia/frank-patterson-cycling-artist-500010568

    Well yes, I get the feeling, believe me. I've ridden exactly 4.9 km
    outside, over the last five months. :-/ I'm trying hard not to become comfortable with that. :-) But that's a different story and not one for
    this forum. Just this much: I've never cycled as far and as high on a
    single day as I have done repeatedly since we retired. Becoming old is
    an obstacle, no question. But you can postpone the consequences of
    ageing, at least for a while.


    Agreed. Components include moving (be that cycling or
    whatever), attitude and probably diet. All of which we
    control with our will; they are not external problems.

    https://nypost.com/2025/01/02/sports/agnes-keleti-oldest-living-olympic-medalist-and-holocaust-survivor-dead-at-103/

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Catrike Ryder@21:1/5 to news51@mystrobl.de on Mon Jan 13 16:40:50 2025
    On Mon, 13 Jan 2025 20:48:36 +0100, Wolfgang Strobl
    <news51@mystrobl.de> wrote:

    Am Mon, 13 Jan 2025 12:27:27 -0500 schrieb Frank Krygowski ><frkrygow@sbcglobal.net>:

    On 1/13/2025 9:57 AM, Wolfgang Strobl wrote:
    Am Sun, 12 Jan 2025 21:05:47 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    On 1/12/2025 3:33 PM, Wolfgang Strobl wrote:
    Am Sat, 11 Jan 2025 19:46:50 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    To me, a big advantage is the ability to _look_ at a mechanical device >>>>>> and _see_ what's wrong....

    That, and the fact I can often affect a repair.

    I prefer devices that don't need repair over their lifetime.

    The weakness I see with that is the assumption that "lifetime" is
    defined as "the amount of time it works." if something stops working,
    its lifetime is over! Throw it out!

    That's far too simplistic.

    It depends. For my purposes, I indeed prefer bicycles that may need
    repairs and modifications over their lifetime, for various reasons. I
    change over my lifetime, so do my bicycles. But there are limits. Want
    it cheap, longlived, lightweight and functional? Choose any two.



    As I said, I hate the Kleenex ethic - "It's no good any more, just throw >>>> it away."

    A strawman isn't getting any more pretty, over time. You won't find many >>> complex products, machines, vehicles or components with an unlimited
    lifetime. Product lifetime has to be planned. There is innovation,
    innovation means change. There are technical limits. So far, I haven't
    heard about bicycle tires that tolerate heavy use over a lifetime of 40
    years, as you ask for. To be precise, I don't know of any that I would
    like to use or that I would risk using.

    I think my Cannondale touring bike qualifies. Of course I've replaced >>consumable items like tires, chains, cogs, brake shoes, handlebar tape
    and occasionally a chainring.

    That way, any bicycle qualifies. There is essentially no part of a
    bicycle that isn't "consumable". If you are lucky, all consumable items
    are consumed at the same time. So you can just buy another bike, call
    it the repaired one and throw out the old one. :-) Given that declaring >something consumed is a rather arbitrary decision, you have quite a lot
    of slack with that.


    I've made some equipment substitutions
    (saddle, bar-end shifters, "aero" brake levers) but the original
    equipment is exceptionally durable.

    That is a tautology. Of course the remaining equipment is exeptionally >durable, otherwise it would have been replaced earlier, for whatever
    reason.


    For one example: The square taper cranks that Tom mocks still work >>perfectly well. I had to replace the original sealed bottom bracket one >>time, but there was no confusion about compatibility (and my cranks did
    not fall off!). The Stronglite roller bearing headset has also lasted >>decades, with one parts replacement. The SunTour rear derailleur is
    still perfect, although I did cheat a bit. When I powder coated our
    bikes, I traded my derailleur for my wife's, figuring hers had many
    fewer miles; but both still work just fine. Wheels are not original
    because I switched from 27" to 700C, but they're 20 years old.

    So why didn't you buy a 40 years old bicycle from somebody who doesn't
    need his bicycle anymore?

    I guess you don't drive a Ford Model T and you don't use an grammophone
    that needs a steel needle for playing shellac records.

    Personally, I am more concerned about how to use a bicycle rather than
    other modes of transport and optimising the bike for that purpose, and I
    am less concerndedabout whether the bike choosen it will last ten,
    twenty or thirty years.

    How long a bicycle lasts depends upon how much it is used and under what >conditions. A bicycle that lasts more than thirty years is most likely a >display piece. That some people like you have the time, space and energy
    to maintain a bicycle much longer than its useful life is under normal >conditions doesn't prove the opposite. That is not an argument against >repairing, but an argument against repairing, whatever the cost. I'm not >talking about money only, here. I mostly miss a sense of proportion.



    Anyway, I see no reason why the wireless shifting of our bikes shouldn't >>> outlive a similar purely mechanical one...

    I guess we'll see, eventually.

    If we don't try, we certainly won't see it. Try to see it the following
    way: _you_ don't have any reason to try a group with wireless shifting
    like the one I built our bikes with, I understand that. So just let
    people like us who experience, like and sometimes need the benefits pay
    the money, try this innovation, and serve as guinea pigs.


    There was a similar problem with our TV, too many separate components. I >>> solved that by using a power strip combined with a separate central
    switch at an easy to reach location. Powering on/off needs two actions: >>> central switch plus a button on the PC, powering off is done via
    keyboard and central switch. That way, all that stuff doesn't consume
    standby power, when not in use.


    I pump the TV sound through our stereo amplifier, which
    has its own remote (whose volume control seems to have stopped working), >>>> the CD/DVD player has a separate remote, etc. etc. If we had a friend
    house sit for us, I'd have to write a manual on how to run the system.

    This can actually be automated quite easily for devices with IR remote
    controls. However, it does require a little programming and soldering
    work.

    About that: A few years ago I got annoyed at the number of remotes. I'd >>read a good review about a programmable universal remote, and bought it.
    I followed the tedious instructions to program it so I could hit one
    button for "Watch TV", another button for "Play CD", another button for >>"Listen to radio" etc.

    It's less than ideal. Part of the problem, I think, is that some of the >>devices use the same signal code as a toggle for "power-on" &
    "power-off", as opposed to a separate code for "On" and "Off." If a
    device is left in the wrong state, things don't work. There was also
    some dimly remembered problem where commands from the remote had to
    arrive at the TV at the proper instant - not too soon, not too late -
    and the program couldn't manage that, despite the nice lady at the 800
    help number trying over and over to cure. (I suppose I could dig back
    into the programming, but I'm not motivated.)

    Some years ago, I helped extending a library that implements both
    reading (decoding) and sending (generating) IR codes, the primary author
    was quite prolific in extending it to any protocol that he got >specificattions and/or samples for. I only wrote a driver part for a not
    yet supported microcontroller, but that was good enough to understand
    some of its workings. Sadly, the project has mostly stalled after 2015.
    The code is still working and small microcontrollers and IR remotes
    don't change that much.

    From your description, I cannot deduce whether a single button press on
    your IR control serves as an "invert the boolean that denotes specific >state" (power on/off, for example), or if it is something else. Some >universal remotes are just simple and stupid recorders, recording and >replaying a bitstream without decoding, perhaps after some signal
    cleanup. Others decode and work from tables.

    From memory, most IR remotes use the NEC protocol, after extracting an >abstract code from the bistream, that code essentially is triple (device >address, command number, modifier), device denoting a specific tv model,
    for example, command number some arbitrary numbering of the keys on the >remote, modifier in this case just a single bit denoting "this is
    comming from a repeating, still pressed keys).

    Usually, the behaviour of a IR remote control is as simple as that.

    So much about the basics. Knowing neither your universal control, nor >anything about the remote control in question, I can't even guess what
    is causing that problem. I could tell you what I would do to analyze
    and perhaps solve it, but that won't help you, because you don't have
    the necessary equipment (and knowledge). And even I have stored away
    most of the stuff I need for such work, in order to get space to build
    and maintain our bikes.




    I've seen many such fans, radiant heaters and the like, where the pull
    chain or drawstring had been lost or damaged. But I have rarely
    misplacted an IR remote, simply because there is no point in moving it
    out of the room where the controlled device is located.

    Your rooms must be much less messy than mine!

    Not really. It's just that the remotes sit on top of the devices that
    they control, when not in use. Most of the time, that is.


    ...when we
    did our first tours with Peugeot bicycles bought in 1978, we were young
    and we mostly rode on the flat. A piece of cake, even with only 2 x 5
    and without indexing.

    As I get older and older, I'm trying to get more comfortable with being
    an old guy who avoids mountains. Maybe some guy in a Frank Patterson >>drawing: >>https://www.worthpoint.com/worthopedia/frank-patterson-cycling-artist-500010568

    Well yes, I get the feeling, believe me. I've ridden exactly 4.9 km
    outside, over the last five months. :-/ I'm trying hard not to become >comfortable with that. :-) But that's a different story and not one for
    this forum. Just this much: I've never cycled as far and as high on a
    single day as I have done repeatedly since we retired. Becoming old is
    an obstacle, no question. But you can postpone the consequences of
    ageing, at least for a while.

    A bicycle frame wears out? Really?

    --
    C'est bon
    Soloman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Catrike Ryder on Mon Jan 13 18:24:35 2025
    On 1/13/2025 3:40 PM, Catrike Ryder wrote:
    On Mon, 13 Jan 2025 20:48:36 +0100, Wolfgang Strobl
    <news51@mystrobl.de> wrote:

    Am Mon, 13 Jan 2025 12:27:27 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    On 1/13/2025 9:57 AM, Wolfgang Strobl wrote:
    Am Sun, 12 Jan 2025 21:05:47 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    On 1/12/2025 3:33 PM, Wolfgang Strobl wrote:
    Am Sat, 11 Jan 2025 19:46:50 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    To me, a big advantage is the ability to _look_ at a mechanical device >>>>>>> and _see_ what's wrong....

    That, and the fact I can often affect a repair.

    I prefer devices that don't need repair over their lifetime.

    The weakness I see with that is the assumption that "lifetime" is
    defined as "the amount of time it works." if something stops working, >>>>> its lifetime is over! Throw it out!

    That's far too simplistic.

    It depends. For my purposes, I indeed prefer bicycles that may need
    repairs and modifications over their lifetime, for various reasons. I
    change over my lifetime, so do my bicycles. But there are limits. Want >>>> it cheap, longlived, lightweight and functional? Choose any two.



    As I said, I hate the Kleenex ethic - "It's no good any more, just throw >>>>> it away."

    A strawman isn't getting any more pretty, over time. You won't find many >>>> complex products, machines, vehicles or components with an unlimited
    lifetime. Product lifetime has to be planned. There is innovation,
    innovation means change. There are technical limits. So far, I haven't >>>> heard about bicycle tires that tolerate heavy use over a lifetime of 40 >>>> years, as you ask for. To be precise, I don't know of any that I would >>>> like to use or that I would risk using.

    I think my Cannondale touring bike qualifies. Of course I've replaced
    consumable items like tires, chains, cogs, brake shoes, handlebar tape
    and occasionally a chainring.

    That way, any bicycle qualifies. There is essentially no part of a
    bicycle that isn't "consumable". If you are lucky, all consumable items
    are consumed at the same time. So you can just buy another bike, call
    it the repaired one and throw out the old one. :-) Given that declaring
    something consumed is a rather arbitrary decision, you have quite a lot
    of slack with that.


    I've made some equipment substitutions
    (saddle, bar-end shifters, "aero" brake levers) but the original
    equipment is exceptionally durable.

    That is a tautology. Of course the remaining equipment is exeptionally
    durable, otherwise it would have been replaced earlier, for whatever
    reason.


    For one example: The square taper cranks that Tom mocks still work
    perfectly well. I had to replace the original sealed bottom bracket one
    time, but there was no confusion about compatibility (and my cranks did
    not fall off!). The Stronglite roller bearing headset has also lasted
    decades, with one parts replacement. The SunTour rear derailleur is
    still perfect, although I did cheat a bit. When I powder coated our
    bikes, I traded my derailleur for my wife's, figuring hers had many
    fewer miles; but both still work just fine. Wheels are not original
    because I switched from 27" to 700C, but they're 20 years old.

    So why didn't you buy a 40 years old bicycle from somebody who doesn't
    need his bicycle anymore?

    I guess you don't drive a Ford Model T and you don't use an grammophone
    that needs a steel needle for playing shellac records.

    Personally, I am more concerned about how to use a bicycle rather than
    other modes of transport and optimising the bike for that purpose, and I
    am less concerndedabout whether the bike choosen it will last ten,
    twenty or thirty years.

    How long a bicycle lasts depends upon how much it is used and under what
    conditions. A bicycle that lasts more than thirty years is most likely a
    display piece. That some people like you have the time, space and energy
    to maintain a bicycle much longer than its useful life is under normal
    conditions doesn't prove the opposite. That is not an argument against
    repairing, but an argument against repairing, whatever the cost. I'm not
    talking about money only, here. I mostly miss a sense of proportion.



    Anyway, I see no reason why the wireless shifting of our bikes shouldn't >>>> outlive a similar purely mechanical one...

    I guess we'll see, eventually.

    If we don't try, we certainly won't see it. Try to see it the following
    way: _you_ don't have any reason to try a group with wireless shifting
    like the one I built our bikes with, I understand that. So just let
    people like us who experience, like and sometimes need the benefits pay
    the money, try this innovation, and serve as guinea pigs.


    There was a similar problem with our TV, too many separate components. I >>>> solved that by using a power strip combined with a separate central
    switch at an easy to reach location. Powering on/off needs two actions: >>>> central switch plus a button on the PC, powering off is done via
    keyboard and central switch. That way, all that stuff doesn't consume >>>> standby power, when not in use.


    I pump the TV sound through our stereo amplifier, which
    has its own remote (whose volume control seems to have stopped working), >>>>> the CD/DVD player has a separate remote, etc. etc. If we had a friend >>>>> house sit for us, I'd have to write a manual on how to run the system. >>>>
    This can actually be automated quite easily for devices with IR remote >>>> controls. However, it does require a little programming and soldering
    work.

    About that: A few years ago I got annoyed at the number of remotes. I'd
    read a good review about a programmable universal remote, and bought it. >>> I followed the tedious instructions to program it so I could hit one
    button for "Watch TV", another button for "Play CD", another button for
    "Listen to radio" etc.

    It's less than ideal. Part of the problem, I think, is that some of the
    devices use the same signal code as a toggle for "power-on" &
    "power-off", as opposed to a separate code for "On" and "Off." If a
    device is left in the wrong state, things don't work. There was also
    some dimly remembered problem where commands from the remote had to
    arrive at the TV at the proper instant - not too soon, not too late -
    and the program couldn't manage that, despite the nice lady at the 800
    help number trying over and over to cure. (I suppose I could dig back
    into the programming, but I'm not motivated.)

    Some years ago, I helped extending a library that implements both
    reading (decoding) and sending (generating) IR codes, the primary author
    was quite prolific in extending it to any protocol that he got
    specificattions and/or samples for. I only wrote a driver part for a not
    yet supported microcontroller, but that was good enough to understand
    some of its workings. Sadly, the project has mostly stalled after 2015.
    The code is still working and small microcontrollers and IR remotes
    don't change that much.

    From your description, I cannot deduce whether a single button press on
    your IR control serves as an "invert the boolean that denotes specific
    state" (power on/off, for example), or if it is something else. Some
    universal remotes are just simple and stupid recorders, recording and
    replaying a bitstream without decoding, perhaps after some signal
    cleanup. Others decode and work from tables.

    From memory, most IR remotes use the NEC protocol, after extracting an
    abstract code from the bistream, that code essentially is triple (device
    address, command number, modifier), device denoting a specific tv model,
    for example, command number some arbitrary numbering of the keys on the
    remote, modifier in this case just a single bit denoting "this is
    comming from a repeating, still pressed keys).

    Usually, the behaviour of a IR remote control is as simple as that.

    So much about the basics. Knowing neither your universal control, nor
    anything about the remote control in question, I can't even guess what
    is causing that problem. I could tell you what I would do to analyze
    and perhaps solve it, but that won't help you, because you don't have
    the necessary equipment (and knowledge). And even I have stored away
    most of the stuff I need for such work, in order to get space to build
    and maintain our bikes.




    I've seen many such fans, radiant heaters and the like, where the pull >>>> chain or drawstring had been lost or damaged. But I have rarely
    misplacted an IR remote, simply because there is no point in moving it >>>> out of the room where the controlled device is located.

    Your rooms must be much less messy than mine!

    Not really. It's just that the remotes sit on top of the devices that
    they control, when not in use. Most of the time, that is.


    ...when we
    did our first tours with Peugeot bicycles bought in 1978, we were young >>>> and we mostly rode on the flat. A piece of cake, even with only 2 x 5
    and without indexing.

    As I get older and older, I'm trying to get more comfortable with being
    an old guy who avoids mountains. Maybe some guy in a Frank Patterson
    drawing:
    https://www.worthpoint.com/worthopedia/frank-patterson-cycling-artist-500010568

    Well yes, I get the feeling, believe me. I've ridden exactly 4.9 km
    outside, over the last five months. :-/ I'm trying hard not to become
    comfortable with that. :-) But that's a different story and not one for
    this forum. Just this much: I've never cycled as far and as high on a
    single day as I have done repeatedly since we retired. Becoming old is
    an obstacle, no question. But you can postpone the consequences of
    ageing, at least for a while.

    A bicycle frame wears out? Really?

    --
    C'est bon
    Soloman

    Some people have very high standards and toss useful things
    when they are merely scuffed or have a small dent. Other
    people have different standards:

    https://www.yellowjersey.org/mitch.html

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Joy Beeson@21:1/5 to news51@mystrobl.de on Mon Jan 13 21:57:59 2025
    On Mon, 13 Jan 2025 21:36:18 +0100, Wolfgang Strobl
    <news51@mystrobl.de> wrote:

    And, to be honest, I don't like the fact
    that you can hardly see the stars anywhere these days either.

    In my previous home, I was very upset that the adjacent high school
    glared such blinding light into my back yard that if I went to the
    garden after dark, I stepped on the rhubarb.

    That was with tennis courts, a windbreak, and a lot that now has a
    house on it in between.

    Heaven knows what they are doing now that LED lights are available.

    --
    Joy Beeson
    joy beeson at centurylink dot net
    http://wlweather.net/PAGEJOY/

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Joy Beeson@21:1/5 to All on Mon Jan 13 23:26:49 2025
    On Sun, 12 Jan 2025 14:13:36 -0800, Jeff Liebermann <jeffl@cruzio.com>
    wrote:

    It's
    really difficult to go back to primitive.

    Partly because there are no sources for primitive things.

    And no infrastrcture. The trailer fridge that had a compartment for a fifty-pound block of ice when you parked somewhere that you couldn't
    plug in was very clever -- but fifty-pound blocks of ice are nowhere
    to be found, and bags of ice cubes don't work at all well.

    The ice compartment was just right for a gallon of milk when you were
    plugged in -- and it was so easy to defrost! Just turn it off, and
    check now and again to see whether all the frost has fallen onto the
    drainboard that catches the melted ice. It stayed plenty cool while defrosting, and you didn't have to move anythng.

    --
    Joy Beeson
    joy beeson at centurylink dot net
    http://wlweather.net/PAGEJOY/

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to frkrygow@sbcglobal.net on Mon Jan 13 21:13:57 2025
    On Mon, 13 Jan 2025 12:58:35 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/13/2025 12:06 PM, Wolfgang Strobl wrote:
    You, as a person, don't need or consume heat. Somewhat simplified, you
    need a certain range of temperatures. In the ideal case, you don't need
    any additional energy, because your body already produces heat. A little
    bit of isolation, perhaps provided by that very blanket, might be
    sufficient.

    And, I suppose, we could do away with all home heating, and just wear
    very heavy clothing all winter. But I don't know of anyone doing that.

    I'm currently doing something similar to save firewood, which costs
    about $500/cord delivered. My bed has multiple layers of blanket and
    a goose feather duvet. During the day at home, I wear a padded jacket
    and fleece lined pants. I'm constantly moving around so I tend to
    warm myself. I usually wear a wool hat indoors to keep my head warm.
    At about sundown, the temperature drops sufficiently to require
    additional heat. The wood burner runs in the evening for about 5 hrs
    until I fall into the bed. If I have guests during the day, I start a
    fire in the wood burner. The reason this works for me is that the
    temperatures are quite mild during most of the winter.

    <https://photos.app.goo.gl/akvXf9MyYNzpYpY77>

    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Roger Merriman@21:1/5 to Frank Krygowski on Tue Jan 14 12:12:13 2025
    Frank Krygowski <frkrygow@sbcglobal.net> wrote:
    On 1/12/2025 12:30 PM, AMuzi wrote:

    https://www.reddit.com/r/cycling/comments/p0vxtq/
    shimano_di2_options_when_battery_dies_mid_trip/

    "One temporary solution would be to find someone that has a Di2, unplug
    their and your rear derailleur and plug in their cable into your
    derailleur. You can then change to the comfortable gear and ride back
    home. "

    From the comments there: "My wireless 12speed di2 dies after roughly 2 weeks."

    Yow!

    And plenty of copies of "Just charge your batteries frequently!"

    No thanks. I don't even want to have to think about headlamp batteries.



    Living in an affluent part of London/Home counties plenty of real word experience of such systems rather than internet tales of exaggeration/fabrication etc!

    People do run them flat but that’s not because they run out in two weeks
    but after years for the hood battery’s, Shimano say up to 3 SRAM say about
    2 years.

    The derailleurs are more months than years, the front mech being more
    demanding than the rear in terms of power consumption but both front/rear Shimano/Sram seem to last months.

    So that the run flat is due to that, ie folks have simply forgotten to
    charge as they don’t routinely charge them.

    The only downside I’ve noted but this is definitely anecdotal than statistically significant is of my fellow Gravel club mates which is
    relatively small group two of which have had derailleurs electronic
    failures, both where covered under warranty, but does bring with it a new
    and extra failure point.

    And these tend to be more expensive than mechanical.

    Roger Merriman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Catrike Ryder@21:1/5 to Roger Merriman on Tue Jan 14 07:45:59 2025
    On 14 Jan 2025 12:22:39 GMT, Roger Merriman <roger@sarlet.com> wrote:

    Frank Krygowski <frkrygow@sbcglobal.net> wrote:
    On 1/13/2025 8:54 AM, zen cycle wrote:
    On 1/12/2025 2:25 PM, Frank Krygowski wrote:
    On 1/12/2025 7:49 AM, zen cycle wrote:
    On 1/11/2025 7:46 PM, Frank Krygowski wrote:

    But individual power outputs from left and right legs? I can't
    imagine much value from that measurement.

    You answered your own question. "You" can't imagine it's useful.
    There are a great many data geeks who find it quite useful.

    Maybe we can take a poll? How many here gather and use data on their
    individual legs' power output?

    I hardly consider this forum to be representative of the cycling
    community at alrge.

    I certainly agree with that! This is a group with no members who were
    not interested in the title "tech". So if these people are not
    measuring individual leg power, I doubt it's very common.

    Its also barely into double digits and its age range is slewed to the more >mature, I suspect Im the youngest here but Im about mid pack for a club >rider ie late 40s/early 50s

    Id also note that due to the age range is a tendency for this effect.

    Noted by Douglas Adams I've come up with a set of rules that describe our >reactions to technologies:
    1. Anything that is in the world when youre born is normal and ordinary
    and is just a natural part of the way the world works.
    2. Anything that's invented between when youre fifteen and thirty-five is >new and exciting and revolutionary and you can probably get a career in it. >3. Anything invented after you're thirty-five is against the natural order
    of things.

    Ie we become less curious as we age in general.

    Our bike club is mostly social, not very competitive. I do have friends
    who use "new" equipment (e.g. aero wheels, 13 cogs, road disc brakes)
    and a few who mention their KOM triumphs. Many of them use Garmin or
    similar devices. But IIRC none have ever discussed power meter readings.
    Maybe some new, younger club members are doing that, but I don't ride
    with those folks.

    Then there's the vast majority of bike riders who will never join a
    club, never look at a bike magazine, never try to ride fast, who just
    ride for fun. No power meters there.

    They will not be selling power meters be that crank or pedal on the efforts >of club riders or racers, just wouldnt be commercially viable!

    Roger Merriman

    I don't know what the quandary is. Some people use them some don't. I
    don't see any right or wrong, good or bad, issues worth arguing about.

    --
    C'est bon
    Soloman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Roger Merriman@21:1/5 to Frank Krygowski on Tue Jan 14 12:22:39 2025
    Frank Krygowski <frkrygow@sbcglobal.net> wrote:
    On 1/13/2025 8:54 AM, zen cycle wrote:
    On 1/12/2025 2:25 PM, Frank Krygowski wrote:
    On 1/12/2025 7:49 AM, zen cycle wrote:
    On 1/11/2025 7:46 PM, Frank Krygowski wrote:

    But individual power outputs from left and right legs? I can't
    imagine much value from that measurement.

    You answered your own question. "You" can't imagine it's useful.
    There are a great many data geeks who find it quite useful.

    Maybe we can take a poll? How many here gather and use data on their
    individual legs' power output?

    I hardly consider this forum to be representative of the cycling
    community at alrge.

    I certainly agree with that! This is a group with no members who were
    not interested in the title "tech". So if these people are not
    measuring individual leg power, I doubt it's very common.

    It’s also barely into double digits and its age range is slewed to the more mature, I suspect I’m the youngest here but I’m about mid pack for a club rider ie late 40’s/early 50’s

    I’d also note that due to the age range is a tendency for this effect.

    Noted by Douglas Adams “I've come up with a set of rules that describe our reactions to technologies:
    1. Anything that is in the world when you’re born is normal and ordinary
    and is just a natural part of the way the world works.
    2. Anything that's invented between when you’re fifteen and thirty-five is new and exciting and revolutionary and you can probably get a career in it.
    3. Anything invented after you're thirty-five is against the natural order
    of things.”

    Ie we become less curious as we age in general.

    Our bike club is mostly social, not very competitive. I do have friends
    who use "new" equipment (e.g. aero wheels, 13 cogs, road disc brakes)
    and a few who mention their KOM triumphs. Many of them use Garmin or
    similar devices. But IIRC none have ever discussed power meter readings. Maybe some new, younger club members are doing that, but I don't ride
    with those folks.

    Then there's the vast majority of bike riders who will never join a
    club, never look at a bike magazine, never try to ride fast, who just
    ride for fun. No power meters there.

    They will not be selling power meters be that crank or pedal on the efforts
    of club riders or racers, just wouldn’t be commercially viable!

    Roger Merriman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Roger Merriman@21:1/5 to Frank Krygowski on Tue Jan 14 12:57:16 2025
    Frank Krygowski <frkrygow@sbcglobal.net> wrote:
    On 1/13/2025 2:48 PM, Wolfgang Strobl wrote:
    Am Mon, 13 Jan 2025 12:27:27 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    As I said, I hate the Kleenex ethic - "It's no good any more, just throw >>>>> it away."

    A strawman isn't getting any more pretty, over time. You won't find many >>>> complex products, machines, vehicles or components with an unlimited
    lifetime. Product lifetime has to be planned. There is innovation,
    innovation means change. There are technical limits. So far, I haven't >>>> heard about bicycle tires that tolerate heavy use over a lifetime of 40 >>>> years, as you ask for. To be precise, I don't know of any that I would >>>> like to use or that I would risk using.

    I think my Cannondale touring bike qualifies. Of course I've replaced
    consumable items like tires, chains, cogs, brake shoes, handlebar tape
    and occasionally a chainring.

    That way, any bicycle qualifies.

    I agree! Or at least, I agree about most bikes. That's one of the things
    I love about bicycling in general, compared to (say) automobiles.

    There is essentially no part of a bicycle that isn't "consumable".

    I disagree. I don't expect to ever wear out the frame, fork, handlebars, stem, seatpost, hubs, pedals, front derailleur, and maybe not the rear derailleur. I may someday wear out the bottle dynamo on that bike (it's decades old) but maybe not.

    Wear out can also mean gets damaged, I’ve replaced 4 rear mech’s due to rock strikes/getting clogged and wrapped around the cassette.

    Likewise folks can and do damage frames, even metal ones, or if steel can corroded to a point that isn’t much to be done and so on.

    The old Commute MTB which is “grandfathers axe” has original frame/stem/bars

    There were items I changed out of preference (like the original downtube shifters) but it wasn't because they were worn out. Those would have
    lasted forever.

    Probably the slower shifts would hide some of lack of snappiness you get
    from older shifters/mech my old school bike who’s front mech’s is definitely not quite right and shifts really quite slowly is about the same
    as the friction shifting, I had on the old steel bike which felt fine.

    Snips

    Roger Merriman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Roger Merriman@21:1/5 to Catrike Ryder on Tue Jan 14 12:41:16 2025
    Catrike Ryder <Soloman@old.bikers.org> wrote:
    On Sun, 12 Jan 2025 19:28:34 -0600, AMuzi <am@yellowjersey.org> wrote:

    On 1/12/2025 7:19 PM, John B. wrote:
    On Sun, 12 Jan 2025 11:08:45 -0600, AMuzi <am@yellowjersey.org> wrote:

    On 1/11/2025 6:46 PM, Frank Krygowski wrote:
    On 1/11/2025 4:11 PM, Wolfgang Strobl wrote:
    Am Fri, 10 Jan 2025 20:01:26 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    On 1/10/2025 5:18 PM, Wolfgang Strobl wrote:
    This makes me reflect on the criticism of electronics in
    bicycles. ... But what about measuring the
    amount of power applied to the pedals, what about
    telling the cyclist
    who balanced he splits the power between left and right, by
    instrumenting the pedal or the bottom bracket?

    Yes, that can be done with enough sensors and
    electronics. But it seems
    like useless information to me. As with much computer
    software, it seems
    like "feature bloat." Why would anybody but a racer care?

    Is there any reason to believe that old people like me
    don't have a need
    to care about training intensity, as a matter of
    principle?  How comes?

    When using my biycles as a middle aged, healthy adult I
    didn't care
    about racing, didn't train and didn't try to find company
    in cycling
    clubs, either. I didn't visit a gym until much later.
    Cycling was simply
    a way to get to work quickly. During vacations it was a more
    entertaining mode of transportation than the car. Fitness,
    strength and
    cardio fitness came as a side effect, slowly and almost
    unnoticed.

    This has changed. My cardio fitness is still better than
    what I know
    about a lot of people half my age. Nevertheless, it
    decreased over the
    years and it takes more and more effort to keep what you
    have not yet
    lost. As it is expected. The ability to measure power is
    helpful in
    finding the point at which intensity is sufficient and
    overload has not
    yet begun. This is even more true in combination with a
    heart rate
    monitor.

    So far, this is only about aging. But there is more.

    Accidents that led to damage to bones, joints and tendons
    in the past
    have consequences to be considered, too. There is a
    tendency to
    compensate weaknesses by bad postures, whithout noticing.
    This is both
    caused by damages and causing damages. This is to be
    avoided.  Measuring
    how power is applied by the feet helps detecting
    differences early,
    avoiding damage.

    Avoiding damage is far from useless.

    But individual power outputs from left and right legs? I
    can't imagine much value from that measurement. In fact, I
    don't believe power measurement is necessary for fitness. I
    judge by feel. It has the advantage of naturally
    compensating for days when I feel stronger or weaker. And
    it's organic!  ;-)

    My experiences with mechanical devices are at least as
    mixed as those
    with electronic devices.  Some very complicated electronic
    devices that
    I bought decades ago still work. The same applies to some
    that I built
    myself a long time ago.  With devices that contained both
    electronic and
    mechanical parts, the fault was more often on the
    mechanical side.

    Anecdotical, I know.

    To me, a big advantage is the ability to _look_ at a
    mechanical device and _see_ what's wrong. That, and the fact
    I can often affect a repair. I hate the Kleenex ethic -
    "It's no good any more, just throw it away."

    Two days ago, my kid asked me to figure out why an
    electric blanket
    wasn't working. The controller refused to turn on. I
    opened it and
    confirmed that it was getting supply voltage. Beyond
    that, the pile of
    dozens of surface mount electronic components was
    incomprehensible to
    me. I suspect Jeff might have been able to diagnose it,
    but not me.
    Here's a photo:
    https://www.flickr.com/photos/16972296@N08/54259119364/
    in/dateposted-public/
    I know electric blankets are old technology. I know they
    functioned well
    for decades with maybe a rheostat and perhaps a couple of
    other
    components. Why add unrepairable complexity?

    Don't know.  We use devices like the blue one in the
    following picture.
    <https://upload.wikimedia.org/wikipedia/commons/4/4b/
    W%C3%A4rmflasche1.jpg>
    and better isolating blankets. No electricity necessary. :-)

    :-) But you imply that _I'm_ the one insufficiently modern?

    A few lights in our house are switched by set of 2 x 3
    inexpensive
    wireless sockets including two remote controls, that I
    bought eleven
    years ago. I've still to replace the batteries. Two of the
    sockets are
    still spares, I have a replacement cell for the remote
    controls stored
    which might live even longer. Standard type, used in
    garage openers and
    burglar alarms, too. Selecting a channel and paring one of
    the four
    buttons of a remote control with one or more of the
    sockets is as easy
    as pie, using a line of dip switches inside those
    devices.  Quite
    similar to pairing switches and derailleur on our bicyles.

    While I avoid having essential functionality in my house
    depend on
    wireless connections, I enjoy having the option, for
    certain use cases,
    though.

    Yuck.

    I bought a new pedestal fan for our back patio last year.
    I've found a fan to be the best defense against mosquitos.
    The previous one still runs, but exposure to sunlight
    destroyed its plastic finger protection screen.

    Anyway, the new one comes with a remote control, roughly the
    size of a snack cracker. Who needs a remote to adjust a fan?
    I rejected ceiling fans with remotes when I was shopping,
    too. It's something to lose, something whose battery will
    die, and something whose electronics will go bad and be
    unrepairable.
    ...

    When I see a 50 year old, perfect looking bicycle, I think:
    That one must have been standing around inside most of the
    time, perhaps
    for a reason.

    My favorite bike is a young 39 years. It looks perfect, in
    part because I had it (and my wife's matching one) stripped
    and powder coated a couple years ago. Too many excellent
    memories to let it go.



    Not only the electronics infestation of consumer products.

    I brought food to make girlfriend's birthday dinner last
    evening and I cook in her kitchen regularly but sometimes
    the modern kitchen stove is a real impediment.

    [break for electronics rant: Her new kitchen stove, as her
    furnace, both run from the large LP tank in the yard. Both
    'improved' designs cannot run without electronic
    start/valves etc so when the power is out there is no heat
    whatsoever and power outages in rural USA are periodic.]

    The new kitchen stove has electronic controls for the
    burners and oven (oven controlled by a touch pad not, a
    physical switch). The burner controls have 4 positions
    only, viz., Off, Low, High, Start. That's really difficult
    for some cooking projects. Resolved by moving the two pans
    off and on the flame, which is Just One More Thing when
    timing two dishes at once.

    Only a designer who has never actually cooked would think of
    that. Works fine for brewing coffee or boiling pasta but
    limited for many projects.

    Gee,,, my mother cooked on a "gas" stove with mechanical controls from
    the regulate at the tank to the knobs on the stove successfully and
    one grandmother cooked on a wood stove, in the winter, un till she
    died, successfully.

    Are these "modern contraptions" really necessary?

    Well, perhaps they are. Are there young woman, today, being taught by
    their mother how to cook on a wood stove? Or how to darn woolen socks,
    or even how to raise the kids?


    Much agreed.

    The kitchen stove I used for decades had simple gas valves
    and a box of wooden matches in a clay dish next to it. No
    complaints.

    I did live for a while on the boat without a microwave, but I'll never
    do that again.

    Indeed though I rarely use a microwave but to defrost stuff!

    I strongly suspect that particularly with domestic appliances that older
    men use of them isn’t a useful metric ie highly likely to be used by women who i suspect would have more experience of such things and different conclusions.
    --
    C'est bon
    Soloman


    Roger Merriman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Roger Merriman@21:1/5 to Wolfgang Strobl on Tue Jan 14 13:03:30 2025
    Wolfgang Strobl <news51@mystrobl.de> wrote:
    Am Mon, 13 Jan 2025 12:58:35 -0500 schrieb Frank Krygowski <frkrygow@sbcglobal.net>:

    On 1/13/2025 12:06 PM, Wolfgang Strobl wrote:
    Am Mon, 13 Jan 2025 11:03:03 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    On 1/13/2025 8:57 AM, zen cycle wrote:
    On 1/13/2025 2:01 AM, Jeff Liebermann wrote:
    On Mon, 13 Jan 2025 13:20:04 +0700, John B. <slocombjb@gmail.com>
    wrote:

    But long before the days of transistors and "circuit boards" my folks >>>>>>> had an electric blanket... used it all the rest of their lives. Is the >>>>>>> modern one better?

    Not really better, but certainly safer.  Electric heaters are all 100% >>>>>> efficient.  Every watt you shove into the heating wires is turned into >>>>>> heat.  There's no way to improve on 100% efficiency.

    Nothing is 100% efficient

    It's certainly true that 100% of the electricity consumed by an electric >>>> blanket becomes heat.

    But that 100% isn't efficiency, it's just a raw conversion rate for
    Electricity to heat in an isolated blanket.

    ISTM a good definition of efficiency is "Desired output divided by
    input." Since the desired output is heat, I think it's 100% efficient.

    "desired output" does some heavy lifting, here. What about the desired input?

    Textbook definitions from a specific context are often quite misleading,
    when used in a different context. Rolf Mantel just gave an example for
    that in <vm38of$1qe41$1@dont-email.me>, there are many more.



    You, as a person, don't need or consume heat. Somewhat simplified, you
    need a certain range of temperatures. In the ideal case, you don't need >>> any additional energy, because your body already produces heat. A little >>> bit of isolation, perhaps provided by that very blanket, might be
    sufficient.

    And, I suppose, we could do away with all home heating, and just wear
    very heavy clothing all winter. But I don't know of anyone doing that.

    Actually, we added some isolation to our house, last year. The reduced
    energy consumption was quite noticeable and better than we expected. No
    heavy clothing necessary. It's an old house, unfortunately some
    regulations prohibit doing more or would make it very expensive. Decades
    ago, a former colleague build a house according to current standards
    from that time, heated by a heat pump and geothermics. Compared to the
    quoted 100 % "efficiency" of your heated blanket, that heating is ~750% efficient. While riding over the land during in 2021ff, I saw not only
    many new collectors on the roofs, mostly photovoltaic now, people were isolating their freestanding houses to such an extent that you would
    hardly believe it if you hadn't seen it.




    <https://www.sciencedirect.com/science/article/abs/pii/S0301421514002638> >>>
    "A brighter future? Quantifying the rebound effect in energy efficient
    lighting"

    <https://www.arquiled.com/en/avoiding-the-rebound-effect-when-transitioning-to-led/>

    "In Portugal, the average amount of light has risen 120% in the last
    five years. It is urgent to prevent the energy savings associated with
    the transition to LED from being offset by unnecessary lighting"

    I don't perceive that to be a serious problem, at least indoors.

    The difference between the expected and actual efficiency is independent
    of whether the lighting is installed indoors or outdoors.

    But there is indeed another and additional problem caused by too much
    light outside: Both astronomers and wildlife are not happy about the increasing light pollution. And, to be honest, I don't like the fact
    that you can hardly see the stars anywhere these days either.

    I find it ranges a lot from dark and clear enough (weather conditions permitting) to see the shape of the galaxy around my folks place, head up
    out of the valley and there is zero visible light but the stars and it’s quite spectacular hence that area has dark sky places for Tourists and so
    on, though that’s generally on lower elevations that are easier to reach.

    But can see a few stars even in london and more so once in the royal parks
    even if can see the orange glow of the light pollution at the edges!

    More commercial areas yes naff all to be seen even have huge lit billboards near work which blots out frankly all but Venus and the moon! And must
    consume many households worth of energy!


    If the
    objective is to expend less energy, it's still been achieved.

    If you are allowed to move the goalpost, the goal is already guaranteed.


    I'm not a fan of overly bright headlamps nor outdoor light pollution,
    but those seem to be separate issues.

    That is true, but obviously no argument. Light pollution is just another detriment of wasting energy by unnecessary lighting.


    I do like/need street lights as my balance system does need light to
    function if on foot, on the bike I have lights to see with or rather I can easily flick from low to high if needed. And choose lights that give a wide beam shape to facilitate my balance systems.

    Roger Merriman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Wolfgang Strobl@21:1/5 to All on Tue Jan 14 14:08:01 2025
    Am Mon, 13 Jan 2025 21:19:51 -0500 schrieb Frank Krygowski <frkrygow@sbcglobal.net>:

    On 1/13/2025 2:48 PM, Wolfgang Strobl wrote:
    Am Mon, 13 Jan 2025 12:27:27 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    As I said, I hate the Kleenex ethic - "It's no good any more, just throw >>>>> it away."

    A strawman isn't getting any more pretty, over time. You won't find many >>>> complex products, machines, vehicles or components with an unlimited
    lifetime. Product lifetime has to be planned. There is innovation,
    innovation means change. There are technical limits. So far, I haven't >>>> heard about bicycle tires that tolerate heavy use over a lifetime of 40 >>>> years, as you ask for. To be precise, I don't know of any that I would >>>> like to use or that I would risk using.

    I think my Cannondale touring bike qualifies. Of course I've replaced
    consumable items like tires, chains, cogs, brake shoes, handlebar tape
    and occasionally a chainring.

    That way, any bicycle qualifies.

    I agree! Or at least, I agree about most bikes. That's one of the things
    I love about bicycling in general, compared to (say) automobiles.

    This is mainly because automobile manufacturing is subject to strict regulations, and rightly so. Some repairs are simply not allowed, others produce a new, different car.


    There is essentially no part of a bicycle that isn't "consumable".

    I disagree. I don't expect to ever wear out the frame, fork, handlebars, >stem, seatpost, hubs, pedals, front derailleur, and maybe not the rear >derailleur. I may someday wear out the bottle dynamo on that bike (it's >decades old) but maybe not.

    It's not about what you expect, but about what is possible and what
    actually happens.

    I had to replace two forks, one handlebar, one stem, a set of pedals, a
    rear derailleur and a whole collection of dynamos, not limited to bottle dynamos. I repaired a broken frame, but I'm sure a workshop would have
    refused such a repair.

    We still own two road bikes from early 2010, having a second life in our
    indoor trainer. An alternative would have been selling those bikes or
    making these a gift. There are a lot of students and other people in my
    home town who use old and somewhat worn race bikes to get around. Our
    campus is somewhat distributed, people neither need nor prefer expensive touring bikes to move between home and school or different lecture
    halls. I still own my second road bike from 1995, a Panasonic PR3000.
    It served as a backup bike, after I bought the aforementioned bike in
    2010. I have no use for that, anymore.

    I could put it on the street for bulky waste collection, like I did with
    the remains of the Peugot from 1978, essentially the frame. The bulky
    waste day is popular with those looking for old furniture, appliances
    and even bicycles. Some repair, others are looking for spare parts, and
    then there are the scrap dealers who generally collect metals. The
    remaining scraps are collected by the waste collection service. In this
    case though, I think the bike is still too good to be used for
    cannibalization. I could sell it at a bike market for loose change, but
    haven't found the time to do that yet. So it's staying in the cellar for
    the time being.

    But outside an emergency, I don't have a reason to use any of these old
    bikes for riding around, anymore. They aged well, because they had good
    use. But different from a living creature, they don't get any better by
    heavy use.


    There were items I changed out of preference (like the original downtube >shifters) but it wasn't because they were worn out. Those would have
    lasted forever.

    From experience, I very much doubt that. These might outlast other
    components though, when used not used during winter on heavyly salted
    roads, like I did with both the Peugeot from '78 and the Panasonic from
    1995. There are a lot of very old bicycles in perfect condition, that
    serve as a collectors item today. Some of them probably never saw a
    road again after the test drive.



    Admittedly, there's an apocryphal tale about someone owning an ancient, >ancient hatchet - sometimes it's been told as George Washington's, or
    Abe Lincoln's, or a great-great-great-grandfather's. Is it the same
    hatchet, even though it's handle was replaced five times and its head
    twice? :-) A more classical version of that question regards the Ship
    of Theseus, here: https://en.wikipedia.org/wiki/Ship_of_Theseus

    Apcryphal is the correct term. Its just a made up story. But it
    exemplifies that extending the meaning of "repair" beyond recognition
    serves no purpose. A repair stretched over much more than the usual
    lifetime of a product is neither a repair nor does it make sense, for
    most products. I'm talking about products one depends on, not about
    collectors items or secondary vehicles.



    I've made some equipment substitutions
    (saddle, bar-end shifters, "aero" brake levers) but the original
    equipment is exceptionally durable.

    That is a tautology. Of course the remaining equipment is exeptionally
    durable, otherwise it would have been replaced earlier, for whatever
    reason.

    Nope, I covered this above. And the shifters and brake levers on my
    "utility" Raleigh, formerly for commuting, now for shopping, are far
    older. Mid 1970s, still working fine.

    I covered that above, see "collectors items".


    So why didn't you buy a 40 years old bicycle from somebody who doesn't
    need his bicycle anymore?

    One reason is in 1986 when I bought this bike, there was no such thing
    as a 40 year old Cannondale touring bike.

    Who said anyting about a Cannondale?

    But ok, there wasn't wireless a SRAM eTap AXS Mullet group in 1995 or
    1978, so I couldn't buy a bike having one, at that time. Neither would
    it have made sense to even try to enhance the Panasonic or the custom
    made bike from 2010 with that group. So I just built a new bike, after I
    didn't find one fitting my requirements, after shopping around.

    So I still don't understand what you are getting at. Even very old bikes
    where new at some point in time. Needs differ and change according to circumstances. You can't expect everyone to have your options and
    preferences.


    That some people like you have the time, space and energy
    to maintain a bicycle much longer than its useful life is under normal
    conditions doesn't prove the opposite. That is not an argument against
    repairing, but an argument against repairing, whatever the cost. I'm not
    talking about money only, here. I mostly miss a sense of proportion.

    I promise to let you know if this bike ever exceeds its "useful life."

    This is arbitrary. Your conditions and circumstances are different from
    mine, mine cant' be generalized, either.


    (We might ask Andrew the age of his fixed gear bike.)

    I don't really care. We both know that fixed gear bikes are exotic. One
    of our sons sent me a snapshot of an unicyclist that he took on the way
    up to Mont Ventoux, last year. So, pray tell, why do you need two
    wheels on a bike for something much simpler, like for example for
    shopping? You do use one of your many bicycles for shopping, right?



    Anyway, I see no reason why the wireless shifting of our bikes shouldn't >>>> outlive a similar purely mechanical one...

    I guess we'll see, eventually.

    If we don't try, we certainly won't see it. Try to see it the following
    way: _you_ don't have any reason to try a group with wireless shifting
    like the one I built our bikes with, I understand that. So just let
    people like us who experience, like and sometimes need the benefits pay
    the money, try this innovation, and serve as guinea pigs.

    Oh, I'm very happy to do that! :-) I've been a Late Adopter of many >technologies. I was rather amazed at myself when I bought the EV.

    Ah, finally! I knew that you would get it. :-)

    Well, we still own a >25 year old car that we bought used, twenty years
    ago. Works well, sort of, because it got well maintained and because we
    rarely used it. Given the fact that an EV would be of no use for us, we currently think about giving up the car, switching to car sharing. We
    haven't done this, yet, because car sharing has its own problems. The
    primary use use case and target group for car sharing are people
    replacing their second car and people who need a car for short trips. We
    fit neither description.


    About that: A few years ago I got annoyed at the number of remotes. I'd
    read a good review about a programmable universal remote, and bought it. >>> I followed the tedious instructions to program it so I could hit one
    button for "Watch TV", another button for "Play CD", another button for
    "Listen to radio" etc.

    It's less than ideal. Part of the problem, I think, is that some of the
    devices use the same signal code as a toggle for "power-on" &
    "power-off", as opposed to a separate code for "On" and "Off." If a
    device is left in the wrong state, things don't work. There was also
    some dimly remembered problem where commands from the remote had to
    arrive at the TV at the proper instant - not too soon, not too late -
    and the program couldn't manage that, despite the nice lady at the 800
    help number trying over and over to cure. (I suppose I could dig back
    into the programming, but I'm not motivated.)...

    Knowing neither your universal control, nor
    anything about the remote control in question, I can't even guess what
    is causing that problem.

    It's a Logitech Harmony 650, bought in 2019. I kept my pages of
    frustrated notes from trying to set it up. They're interesting to read >through. It doesn't matter, though. I'm getting along with it now,
    partly be ignoring what's supposed to be a lot of its capabilities.

    Unfortunately, I don't know anything about Logitechs Harmony series. I
    heard about it, but that's it. Existing open source software for it is
    mostly about how to program those remotes using data grabbed from
    Logitechs website. A package still existing in Debian derivates hasn't
    updated by the original developer since 2015, the website itself doesn't
    exist anymore, for some years.

    <https://web.archive.org/web/20211204023258/https://www.phildev.net/harmony/>

    A short search didn't find any technical info about the internals of
    Logitechs IR products, so as a I said, I can't even guess where it may
    stumble and for what specific reasons.

    Given that recording and and replaying one or more IR remote controls as
    a abstract sequence of commands is quite simple using available open
    source software, given some experience with MC hardware and software, I
    didn't care how Logitech did that, so far. But I find it strange that
    it doesn't just work.

    A large collection of codes of existing products at Logitech's
    database/website is the only added value this product has, IMO. As far
    as I understood it, you can or must update the set of products the
    Logitech remote understands about, no need to record it using the
    universal control. Perhaps that recording is just a primitive fallback function that may or may not work correctly?

    Debian derivates have a package called concordance, probably based onthe aforementioned software, that is able to do that updating/programming on
    Linux, according to the description. Some sources are on github, Other
    people may know more about it, I don't.


    --
    Thank you for observing all safety precautions

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Catrike Ryder@21:1/5 to AMuzi on Tue Jan 14 08:46:09 2025
    On Mon, 13 Jan 2025 18:24:35 -0600, AMuzi <am@yellowjersey.org> wrote:

    On 1/13/2025 3:40 PM, Catrike Ryder wrote:
    On Mon, 13 Jan 2025 20:48:36 +0100, Wolfgang Strobl
    <news51@mystrobl.de> wrote:

    Am Mon, 13 Jan 2025 12:27:27 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    On 1/13/2025 9:57 AM, Wolfgang Strobl wrote:
    Am Sun, 12 Jan 2025 21:05:47 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    On 1/12/2025 3:33 PM, Wolfgang Strobl wrote:
    Am Sat, 11 Jan 2025 19:46:50 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    To me, a big advantage is the ability to _look_ at a mechanical device >>>>>>>> and _see_ what's wrong....

    That, and the fact I can often affect a repair.

    I prefer devices that don't need repair over their lifetime.

    The weakness I see with that is the assumption that "lifetime" is
    defined as "the amount of time it works." if something stops working, >>>>>> its lifetime is over! Throw it out!

    That's far too simplistic.

    It depends. For my purposes, I indeed prefer bicycles that may need >>>>> repairs and modifications over their lifetime, for various reasons. I >>>>> change over my lifetime, so do my bicycles. But there are limits. Want >>>>> it cheap, longlived, lightweight and functional? Choose any two.



    As I said, I hate the Kleenex ethic - "It's no good any more, just throw >>>>>> it away."

    A strawman isn't getting any more pretty, over time. You won't find many >>>>> complex products, machines, vehicles or components with an unlimited >>>>> lifetime. Product lifetime has to be planned. There is innovation,
    innovation means change. There are technical limits. So far, I haven't >>>>> heard about bicycle tires that tolerate heavy use over a lifetime of 40 >>>>> years, as you ask for. To be precise, I don't know of any that I would >>>>> like to use or that I would risk using.

    I think my Cannondale touring bike qualifies. Of course I've replaced
    consumable items like tires, chains, cogs, brake shoes, handlebar tape >>>> and occasionally a chainring.

    That way, any bicycle qualifies. There is essentially no part of a
    bicycle that isn't "consumable". If you are lucky, all consumable items >>> are consumed at the same time. So you can just buy another bike, call
    it the repaired one and throw out the old one. :-) Given that declaring >>> something consumed is a rather arbitrary decision, you have quite a lot
    of slack with that.


    I've made some equipment substitutions
    (saddle, bar-end shifters, "aero" brake levers) but the original
    equipment is exceptionally durable.

    That is a tautology. Of course the remaining equipment is exeptionally
    durable, otherwise it would have been replaced earlier, for whatever
    reason.


    For one example: The square taper cranks that Tom mocks still work
    perfectly well. I had to replace the original sealed bottom bracket one >>>> time, but there was no confusion about compatibility (and my cranks did >>>> not fall off!). The Stronglite roller bearing headset has also lasted
    decades, with one parts replacement. The SunTour rear derailleur is
    still perfect, although I did cheat a bit. When I powder coated our
    bikes, I traded my derailleur for my wife's, figuring hers had many
    fewer miles; but both still work just fine. Wheels are not original
    because I switched from 27" to 700C, but they're 20 years old.

    So why didn't you buy a 40 years old bicycle from somebody who doesn't
    need his bicycle anymore?

    I guess you don't drive a Ford Model T and you don't use an grammophone
    that needs a steel needle for playing shellac records.

    Personally, I am more concerned about how to use a bicycle rather than
    other modes of transport and optimising the bike for that purpose, and I >>> am less concerndedabout whether the bike choosen it will last ten,
    twenty or thirty years.

    How long a bicycle lasts depends upon how much it is used and under what >>> conditions. A bicycle that lasts more than thirty years is most likely a >>> display piece. That some people like you have the time, space and energy >>> to maintain a bicycle much longer than its useful life is under normal
    conditions doesn't prove the opposite. That is not an argument against
    repairing, but an argument against repairing, whatever the cost. I'm not >>> talking about money only, here. I mostly miss a sense of proportion.



    Anyway, I see no reason why the wireless shifting of our bikes shouldn't >>>>> outlive a similar purely mechanical one...

    I guess we'll see, eventually.

    If we don't try, we certainly won't see it. Try to see it the following
    way: _you_ don't have any reason to try a group with wireless shifting
    like the one I built our bikes with, I understand that. So just let
    people like us who experience, like and sometimes need the benefits pay
    the money, try this innovation, and serve as guinea pigs.


    There was a similar problem with our TV, too many separate components. I >>>>> solved that by using a power strip combined with a separate central
    switch at an easy to reach location. Powering on/off needs two actions: >>>>> central switch plus a button on the PC, powering off is done via
    keyboard and central switch. That way, all that stuff doesn't consume >>>>> standby power, when not in use.


    I pump the TV sound through our stereo amplifier, which
    has its own remote (whose volume control seems to have stopped working), >>>>>> the CD/DVD player has a separate remote, etc. etc. If we had a friend >>>>>> house sit for us, I'd have to write a manual on how to run the system. >>>>>
    This can actually be automated quite easily for devices with IR remote >>>>> controls. However, it does require a little programming and soldering >>>>> work.

    About that: A few years ago I got annoyed at the number of remotes. I'd >>>> read a good review about a programmable universal remote, and bought it. >>>> I followed the tedious instructions to program it so I could hit one
    button for "Watch TV", another button for "Play CD", another button for >>>> "Listen to radio" etc.

    It's less than ideal. Part of the problem, I think, is that some of the >>>> devices use the same signal code as a toggle for "power-on" &
    "power-off", as opposed to a separate code for "On" and "Off." If a
    device is left in the wrong state, things don't work. There was also
    some dimly remembered problem where commands from the remote had to
    arrive at the TV at the proper instant - not too soon, not too late -
    and the program couldn't manage that, despite the nice lady at the 800 >>>> help number trying over and over to cure. (I suppose I could dig back
    into the programming, but I'm not motivated.)

    Some years ago, I helped extending a library that implements both
    reading (decoding) and sending (generating) IR codes, the primary author >>> was quite prolific in extending it to any protocol that he got
    specificattions and/or samples for. I only wrote a driver part for a not >>> yet supported microcontroller, but that was good enough to understand
    some of its workings. Sadly, the project has mostly stalled after 2015. >>> The code is still working and small microcontrollers and IR remotes
    don't change that much.

    From your description, I cannot deduce whether a single button press on
    your IR control serves as an "invert the boolean that denotes specific
    state" (power on/off, for example), or if it is something else. Some
    universal remotes are just simple and stupid recorders, recording and
    replaying a bitstream without decoding, perhaps after some signal
    cleanup. Others decode and work from tables.

    From memory, most IR remotes use the NEC protocol, after extracting an
    abstract code from the bistream, that code essentially is triple (device >>> address, command number, modifier), device denoting a specific tv model, >>> for example, command number some arbitrary numbering of the keys on the
    remote, modifier in this case just a single bit denoting "this is
    comming from a repeating, still pressed keys).

    Usually, the behaviour of a IR remote control is as simple as that.

    So much about the basics. Knowing neither your universal control, nor
    anything about the remote control in question, I can't even guess what
    is causing that problem. I could tell you what I would do to analyze
    and perhaps solve it, but that won't help you, because you don't have
    the necessary equipment (and knowledge). And even I have stored away
    most of the stuff I need for such work, in order to get space to build
    and maintain our bikes.




    I've seen many such fans, radiant heaters and the like, where the pull >>>>> chain or drawstring had been lost or damaged. But I have rarely
    misplacted an IR remote, simply because there is no point in moving it >>>>> out of the room where the controlled device is located.

    Your rooms must be much less messy than mine!

    Not really. It's just that the remotes sit on top of the devices that
    they control, when not in use. Most of the time, that is.


    ...when we
    did our first tours with Peugeot bicycles bought in 1978, we were young >>>>> and we mostly rode on the flat. A piece of cake, even with only 2 x 5 >>>>> and without indexing.

    As I get older and older, I'm trying to get more comfortable with being >>>> an old guy who avoids mountains. Maybe some guy in a Frank Patterson
    drawing:
    https://www.worthpoint.com/worthopedia/frank-patterson-cycling-artist-500010568

    Well yes, I get the feeling, believe me. I've ridden exactly 4.9 km
    outside, over the last five months. :-/ I'm trying hard not to become
    comfortable with that. :-) But that's a different story and not one for
    this forum. Just this much: I've never cycled as far and as high on a
    single day as I have done repeatedly since we retired. Becoming old is
    an obstacle, no question. But you can postpone the consequences of
    ageing, at least for a while.

    A bicycle frame wears out? Really?

    --
    C'est bon
    Soloman

    Some people have very high standards and toss useful things
    when they are merely scuffed or have a small dent. Other
    people have different standards:

    https://www.yellowjersey.org/mitch.html

    The Catrike frame is aluminum. Everything is easily replacable and
    almost all of it has been replaced. Actually, I had to replace the
    frame because of hairline cracks.

    --
    C'est bon
    Soloman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Catrike Ryder@21:1/5 to Roger Merriman on Tue Jan 14 09:03:30 2025
    On 14 Jan 2025 12:41:16 GMT, Roger Merriman <roger@sarlet.com> wrote:

    Catrike Ryder <Soloman@old.bikers.org> wrote:
    On Sun, 12 Jan 2025 19:28:34 -0600, AMuzi <am@yellowjersey.org> wrote:

    On 1/12/2025 7:19 PM, John B. wrote:
    On Sun, 12 Jan 2025 11:08:45 -0600, AMuzi <am@yellowjersey.org> wrote: >>>>
    On 1/11/2025 6:46 PM, Frank Krygowski wrote:
    On 1/11/2025 4:11 PM, Wolfgang Strobl wrote:
    Am Fri, 10 Jan 2025 20:01:26 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    On 1/10/2025 5:18 PM, Wolfgang Strobl wrote:
    This makes me reflect on the criticism of electronics in
    bicycles. ... But what about measuring the
    amount of power applied to the pedals, what about
    telling the cyclist
    who balanced he splits the power between left and right, by
    instrumenting the pedal or the bottom bracket?

    Yes, that can be done with enough sensors and
    electronics. But it seems
    like useless information to me. As with much computer
    software, it seems
    like "feature bloat." Why would anybody but a racer care?

    Is there any reason to believe that old people like me
    don't have a need
    to care about training intensity, as a matter of
    principle? How comes?

    When using my biycles as a middle aged, healthy adult I
    didn't care
    about racing, didn't train and didn't try to find company
    in cycling
    clubs, either. I didn't visit a gym until much later.
    Cycling was simply
    a way to get to work quickly. During vacations it was a more
    entertaining mode of transportation than the car. Fitness,
    strength and
    cardio fitness came as a side effect, slowly and almost
    unnoticed.

    This has changed. My cardio fitness is still better than
    what I know
    about a lot of people half my age. Nevertheless, it
    decreased over the
    years and it takes more and more effort to keep what you
    have not yet
    lost. As it is expected. The ability to measure power is
    helpful in
    finding the point at which intensity is sufficient and
    overload has not
    yet begun. This is even more true in combination with a
    heart rate
    monitor.

    So far, this is only about aging. But there is more.

    Accidents that led to damage to bones, joints and tendons
    in the past
    have consequences to be considered, too. There is a
    tendency to
    compensate weaknesses by bad postures, whithout noticing.
    This is both
    caused by damages and causing damages. This is to be
    avoided. Measuring
    how power is applied by the feet helps detecting
    differences early,
    avoiding damage.

    Avoiding damage is far from useless.

    But individual power outputs from left and right legs? I
    can't imagine much value from that measurement. In fact, I
    don't believe power measurement is necessary for fitness. I
    judge by feel. It has the advantage of naturally
    compensating for days when I feel stronger or weaker. And
    it's organic! ;-)

    My experiences with mechanical devices are at least as
    mixed as those
    with electronic devices. Some very complicated electronic
    devices that
    I bought decades ago still work. The same applies to some
    that I built
    myself a long time ago. With devices that contained both
    electronic and
    mechanical parts, the fault was more often on the
    mechanical side.

    Anecdotical, I know.

    To me, a big advantage is the ability to _look_ at a
    mechanical device and _see_ what's wrong. That, and the fact
    I can often affect a repair. I hate the Kleenex ethic -
    "It's no good any more, just throw it away."

    Two days ago, my kid asked me to figure out why an
    electric blanket
    wasn't working. The controller refused to turn on. I
    opened it and
    confirmed that it was getting supply voltage. Beyond
    that, the pile of
    dozens of surface mount electronic components was
    incomprehensible to
    me. I suspect Jeff might have been able to diagnose it,
    but not me.
    Here's a photo:
    https://www.flickr.com/photos/16972296@N08/54259119364/
    in/dateposted-public/
    I know electric blankets are old technology. I know they
    functioned well
    for decades with maybe a rheostat and perhaps a couple of
    other
    components. Why add unrepairable complexity?

    Don't know. We use devices like the blue one in the
    following picture.
    <https://upload.wikimedia.org/wikipedia/commons/4/4b/
    W%C3%A4rmflasche1.jpg>
    and better isolating blankets. No electricity necessary. :-)

    :-) But you imply that _I'm_ the one insufficiently modern?

    A few lights in our house are switched by set of 2 x 3
    inexpensive
    wireless sockets including two remote controls, that I
    bought eleven
    years ago. I've still to replace the batteries. Two of the
    sockets are
    still spares, I have a replacement cell for the remote
    controls stored
    which might live even longer. Standard type, used in
    garage openers and
    burglar alarms, too. Selecting a channel and paring one of
    the four
    buttons of a remote control with one or more of the
    sockets is as easy
    as pie, using a line of dip switches inside those
    devices. Quite
    similar to pairing switches and derailleur on our bicyles.

    While I avoid having essential functionality in my house
    depend on
    wireless connections, I enjoy having the option, for
    certain use cases,
    though.

    Yuck.

    I bought a new pedestal fan for our back patio last year.
    I've found a fan to be the best defense against mosquitos.
    The previous one still runs, but exposure to sunlight
    destroyed its plastic finger protection screen.

    Anyway, the new one comes with a remote control, roughly the
    size of a snack cracker. Who needs a remote to adjust a fan?
    I rejected ceiling fans with remotes when I was shopping,
    too. It's something to lose, something whose battery will
    die, and something whose electronics will go bad and be
    unrepairable.
    ...

    When I see a 50 year old, perfect looking bicycle, I think:
    That one must have been standing around inside most of the
    time, perhaps
    for a reason.

    My favorite bike is a young 39 years. It looks perfect, in
    part because I had it (and my wife's matching one) stripped
    and powder coated a couple years ago. Too many excellent
    memories to let it go.



    Not only the electronics infestation of consumer products.

    I brought food to make girlfriend's birthday dinner last
    evening and I cook in her kitchen regularly but sometimes
    the modern kitchen stove is a real impediment.

    [break for electronics rant: Her new kitchen stove, as her
    furnace, both run from the large LP tank in the yard. Both
    'improved' designs cannot run without electronic
    start/valves etc so when the power is out there is no heat
    whatsoever and power outages in rural USA are periodic.]

    The new kitchen stove has electronic controls for the
    burners and oven (oven controlled by a touch pad not, a
    physical switch). The burner controls have 4 positions
    only, viz., Off, Low, High, Start. That's really difficult
    for some cooking projects. Resolved by moving the two pans
    off and on the flame, which is Just One More Thing when
    timing two dishes at once.

    Only a designer who has never actually cooked would think of
    that. Works fine for brewing coffee or boiling pasta but
    limited for many projects.

    Gee,,, my mother cooked on a "gas" stove with mechanical controls from >>>> the regulate at the tank to the knobs on the stove successfully and
    one grandmother cooked on a wood stove, in the winter, un till she
    died, successfully.

    Are these "modern contraptions" really necessary?

    Well, perhaps they are. Are there young woman, today, being taught by
    their mother how to cook on a wood stove? Or how to darn woolen socks, >>>> or even how to raise the kids?


    Much agreed.

    The kitchen stove I used for decades had simple gas valves
    and a box of wooden matches in a clay dish next to it. No
    complaints.

    I did live for a while on the boat without a microwave, but I'll never
    do that again.

    Indeed though I rarely use a microwave but to defrost stuff!

    I strongly suspect that particularly with domestic appliances that older
    men use of them isnt a useful metric ie highly likely to be used by women >who i suspect would have more experience of such things and different >conclusions.
    --
    C'est bon
    Soloman


    Roger Merriman


    I do defrost with the microwave, but we use it for many other things.
    My wife is a microwave popcorn fanatic. I use it for my morning
    oatmeal and to reheat my coffee. I'm very big on vegetables and they
    go from frozen to edible via the microwave. Leftovers must be reheated
    and potatoes must be made edible.

    --
    C'est bon
    Soloman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Jeff Liebermann on Tue Jan 14 08:26:59 2025
    On 1/13/2025 11:13 PM, Jeff Liebermann wrote:
    On Mon, 13 Jan 2025 12:58:35 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/13/2025 12:06 PM, Wolfgang Strobl wrote:
    You, as a person, don't need or consume heat. Somewhat simplified, you
    need a certain range of temperatures. In the ideal case, you don't need >>> any additional energy, because your body already produces heat. A little >>> bit of isolation, perhaps provided by that very blanket, might be
    sufficient.

    And, I suppose, we could do away with all home heating, and just wear
    very heavy clothing all winter. But I don't know of anyone doing that.

    I'm currently doing something similar to save firewood, which costs
    about $500/cord delivered. My bed has multiple layers of blanket and
    a goose feather duvet. During the day at home, I wear a padded jacket
    and fleece lined pants. I'm constantly moving around so I tend to
    warm myself. I usually wear a wool hat indoors to keep my head warm.
    At about sundown, the temperature drops sufficiently to require
    additional heat. The wood burner runs in the evening for about 5 hrs
    until I fall into the bed. If I have guests during the day, I start a
    fire in the wood burner. The reason this works for me is that the temperatures are quite mild during most of the winter.

    <https://photos.app.goo.gl/akvXf9MyYNzpYpY77>

    Right. We all make our arrangements based on the ambient
    conditions, personal evaluation of comfort and cost. Your
    system works for you, but here it's zero or "like minus 16F"
    while shoveling outside in the wind.

    My very close friend, who's 82, lives alone and her two year
    old feature-laden wonder furnace quit Sunday evening at
    three degrees. The tech said it was a 'transducer' and the
    part is 'on order'.

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to John B. on Tue Jan 14 08:28:06 2025
    On 1/14/2025 3:04 AM, John B. wrote:
    On Mon, 13 Jan 2025 21:13:57 -0800, Jeff Liebermann <jeffl@cruzio.com>
    wrote:

    On Mon, 13 Jan 2025 12:58:35 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/13/2025 12:06 PM, Wolfgang Strobl wrote:
    You, as a person, don't need or consume heat. Somewhat simplified, you >>>> need a certain range of temperatures. In the ideal case, you don't need >>>> any additional energy, because your body already produces heat. A little >>>> bit of isolation, perhaps provided by that very blanket, might be
    sufficient.

    And, I suppose, we could do away with all home heating, and just wear
    very heavy clothing all winter. But I don't know of anyone doing that.

    I'm currently doing something similar to save firewood, which costs
    about $500/cord delivered. My bed has multiple layers of blanket and
    a goose feather duvet. During the day at home, I wear a padded jacket
    and fleece lined pants. I'm constantly moving around so I tend to
    warm myself. I usually wear a wool hat indoors to keep my head warm.
    At about sundown, the temperature drops sufficiently to require
    additional heat. The wood burner runs in the evening for about 5 hrs
    until I fall into the bed. If I have guests during the day, I start a
    fire in the wood burner. The reason this works for me is that the
    temperatures are quite mild during most of the winter.

    <https://photos.app.goo.gl/akvXf9MyYNzpYpY77>


    I don't know whether you know this but when using a stove to heat a
    room put the stove in the opposite side of the room from the chimney.
    Then run the stovepipe from the stove straight up from the stove to
    the ceiling and then along the ceiling, using mounts to separate the
    stove pipe and ceiling, of course, across the room to the chimney.

    That way you get more heat from the same amount of wood.



    And more flue to clean. Which is more important than it may
    at first seem.

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Catrike Ryder@21:1/5 to AMuzi on Tue Jan 14 09:33:02 2025
    On Tue, 14 Jan 2025 08:28:06 -0600, AMuzi <am@yellowjersey.org> wrote:

    On 1/14/2025 3:04 AM, John B. wrote:
    On Mon, 13 Jan 2025 21:13:57 -0800, Jeff Liebermann <jeffl@cruzio.com>
    wrote:

    On Mon, 13 Jan 2025 12:58:35 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/13/2025 12:06 PM, Wolfgang Strobl wrote:
    You, as a person, don't need or consume heat. Somewhat simplified, you >>>>> need a certain range of temperatures. In the ideal case, you don't need >>>>> any additional energy, because your body already produces heat. A little >>>>> bit of isolation, perhaps provided by that very blanket, might be
    sufficient.

    And, I suppose, we could do away with all home heating, and just wear
    very heavy clothing all winter. But I don't know of anyone doing that.

    I'm currently doing something similar to save firewood, which costs
    about $500/cord delivered. My bed has multiple layers of blanket and
    a goose feather duvet. During the day at home, I wear a padded jacket
    and fleece lined pants. I'm constantly moving around so I tend to
    warm myself. I usually wear a wool hat indoors to keep my head warm.
    At about sundown, the temperature drops sufficiently to require
    additional heat. The wood burner runs in the evening for about 5 hrs
    until I fall into the bed. If I have guests during the day, I start a
    fire in the wood burner. The reason this works for me is that the
    temperatures are quite mild during most of the winter.

    <https://photos.app.goo.gl/akvXf9MyYNzpYpY77>


    I don't know whether you know this but when using a stove to heat a
    room put the stove in the opposite side of the room from the chimney.
    Then run the stovepipe from the stove straight up from the stove to
    the ceiling and then along the ceiling, using mounts to separate the
    stove pipe and ceiling, of course, across the room to the chimney.

    That way you get more heat from the same amount of wood.



    And more flue to clean. Which is more important than it may
    at first seem.

    Not to mention draft issues.

    --
    C'est bon
    Soloman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to All on Tue Jan 14 07:40:45 2025
    On Tue, 14 Jan 2025 16:04:11 +0700, John B. <slocombjb@gmail.com>
    wrote:

    I don't know whether you know this but when using a stove to heat a
    room put the stove in the opposite side of the room from the chimney.
    Then run the stovepipe from the stove straight up from the stove to
    the ceiling and then along the ceiling, using mounts to separate the
    stove pipe and ceiling, of course, across the room to the chimney.

    That way you get more heat from the same amount of wood.

    Bad idea. When I bought the house in 1974(?), that's roughly what the
    previous owner had done. I couldn't find a photo, but it was a rather
    large "pot belly" stove that leaked air from every seam. It was
    located at one end of the living room with about a 10 ft horizontal
    flue pipe to where it connected to the a 5 ft vertical insulated
    external pipe (Metalbestos) that acted as a chimney pipe. It didn't
    work for a variety of reasons. The big one was that for the "draw" or
    suction to be able to move sufficient hot air up the chimney pipe, the
    inside of the chimney pipe had to be hot. In order to heat the pipe,
    the hot air from the stove has to rise, which is rather awkward with a
    10 ft horizontal uninsulated pipe with nowhere for the hot air to
    rise. Getting a fire started was possible, but difficult and VERY
    smoky. Cleaning the ash out of the horizontal section was difficult
    because the previous owner had neglected to install a "T" where the
    pipe changed from horizontal to vertical. Even if there was a "T", it
    wouldn't have worked because rising hot air does not like making sharp
    turns.

    In other words, it didn't work and horizontal flue pipe are a really
    bad idea. I had to remodel the living room, move the stairs going
    through the floor, replace the pot belly with an airtight stove, and
    extend the external chimney pipe to 12 ft, in order to fix the
    problems. Although I don't completely understand how it works, I've
    found that the "efficiency" of the system is controlled by how much
    time the hot gasses remain in the stove and whether the wood is
    sufficiently dry (under 10% moisture).


    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to cyclintom on Tue Jan 14 10:23:18 2025
    On 1/14/2025 10:11 AM, cyclintom wrote:
    On Sun Jan 5 05:49:38 2025 zen cycle wrote:

    I was actually thinking your college years would have predated the
    HP-35, but wasn't sure and was too lazy to look it up.

    I remember the day my father brought home a Bowmar Brain.

    http://www.vintagecalculators.com/html/bowmar_calculators.html

    I was a freshman in high school, would have been 1975.




    That's interesting. You've been telling us that you race. That would put you racing at 66 years old. Really competitive.

    You don't know that. I don't either.

    But bicycle races are age-sorted, so racing against other 60
    `70 year olds is as competitive as the 'open' 18 to 40 class.

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Radey Shouman@21:1/5 to Wolfgang Strobl on Tue Jan 14 11:51:17 2025
    Wolfgang Strobl <news51@mystrobl.de> writes:

    Am Mon, 13 Jan 2025 12:27:27 -0500 schrieb Frank Krygowski <frkrygow@sbcglobal.net>:

    On 1/13/2025 9:57 AM, Wolfgang Strobl wrote:
    Am Sun, 12 Jan 2025 21:05:47 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    On 1/12/2025 3:33 PM, Wolfgang Strobl wrote:
    Am Sat, 11 Jan 2025 19:46:50 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    To me, a big advantage is the ability to _look_ at a mechanical device >>>>>> and _see_ what's wrong....

    That, and the fact I can often affect a repair.

    I prefer devices that don't need repair over their lifetime.

    The weakness I see with that is the assumption that "lifetime" is
    defined as "the amount of time it works." if something stops working,
    its lifetime is over! Throw it out!

    That's far too simplistic.

    It depends. For my purposes, I indeed prefer bicycles that may need
    repairs and modifications over their lifetime, for various reasons. I
    change over my lifetime, so do my bicycles. But there are limits. Want
    it cheap, longlived, lightweight and functional? Choose any two.



    As I said, I hate the Kleenex ethic - "It's no good any more, just throw >>>> it away."

    A strawman isn't getting any more pretty, over time. You won't find many >>> complex products, machines, vehicles or components with an unlimited
    lifetime. Product lifetime has to be planned. There is innovation,
    innovation means change. There are technical limits. So far, I haven't
    heard about bicycle tires that tolerate heavy use over a lifetime of 40
    years, as you ask for. To be precise, I don't know of any that I would
    like to use or that I would risk using.

    I think my Cannondale touring bike qualifies. Of course I've replaced >>consumable items like tires, chains, cogs, brake shoes, handlebar tape
    and occasionally a chainring.

    That way, any bicycle qualifies. There is essentially no part of a
    bicycle that isn't "consumable". If you are lucky, all consumable items
    are consumed at the same time. So you can just buy another bike, call
    it the repaired one and throw out the old one. :-) Given that declaring something consumed is a rather arbitrary decision, you have quite a lot
    of slack with that.


    I've made some equipment substitutions
    (saddle, bar-end shifters, "aero" brake levers) but the original
    equipment is exceptionally durable.

    That is a tautology. Of course the remaining equipment is exeptionally durable, otherwise it would have been replaced earlier, for whatever
    reason.


    For one example: The square taper cranks that Tom mocks still work >>perfectly well. I had to replace the original sealed bottom bracket one >>time, but there was no confusion about compatibility (and my cranks did
    not fall off!). The Stronglite roller bearing headset has also lasted >>decades, with one parts replacement. The SunTour rear derailleur is
    still perfect, although I did cheat a bit. When I powder coated our
    bikes, I traded my derailleur for my wife's, figuring hers had many
    fewer miles; but both still work just fine. Wheels are not original
    because I switched from 27" to 700C, but they're 20 years old.

    So why didn't you buy a 40 years old bicycle from somebody who doesn't
    need his bicycle anymore?

    I guess you don't drive a Ford Model T and you don't use an grammophone
    that needs a steel needle for playing shellac records.

    Personally, I am more concerned about how to use a bicycle rather than
    other modes of transport and optimising the bike for that purpose, and I
    am less concerndedabout whether the bike choosen it will last ten,
    twenty or thirty years.

    How long a bicycle lasts depends upon how much it is used and under what conditions. A bicycle that lasts more than thirty years is most likely a display piece. That some people like you have the time, space and energy
    to maintain a bicycle much longer than its useful life is under normal conditions doesn't prove the opposite. That is not an argument against repairing, but an argument against repairing, whatever the cost. I'm not talking about money only, here. I mostly miss a sense of proportion.



    Anyway, I see no reason why the wireless shifting of our bikes shouldn't >>> outlive a similar purely mechanical one...

    I guess we'll see, eventually.

    If we don't try, we certainly won't see it. Try to see it the following
    way: _you_ don't have any reason to try a group with wireless shifting
    like the one I built our bikes with, I understand that. So just let
    people like us who experience, like and sometimes need the benefits pay
    the money, try this innovation, and serve as guinea pigs.


    There was a similar problem with our TV, too many separate components. I >>> solved that by using a power strip combined with a separate central
    switch at an easy to reach location. Powering on/off needs two actions: >>> central switch plus a button on the PC, powering off is done via
    keyboard and central switch. That way, all that stuff doesn't consume
    standby power, when not in use.


    I pump the TV sound through our stereo amplifier, which
    has its own remote (whose volume control seems to have stopped working), >>>> the CD/DVD player has a separate remote, etc. etc. If we had a friend
    house sit for us, I'd have to write a manual on how to run the system.

    This can actually be automated quite easily for devices with IR remote
    controls. However, it does require a little programming and soldering
    work.

    About that: A few years ago I got annoyed at the number of remotes. I'd >>read a good review about a programmable universal remote, and bought it.
    I followed the tedious instructions to program it so I could hit one
    button for "Watch TV", another button for "Play CD", another button for >>"Listen to radio" etc.

    It's less than ideal. Part of the problem, I think, is that some of the >>devices use the same signal code as a toggle for "power-on" &
    "power-off", as opposed to a separate code for "On" and "Off." If a
    device is left in the wrong state, things don't work. There was also
    some dimly remembered problem where commands from the remote had to
    arrive at the TV at the proper instant - not too soon, not too late -
    and the program couldn't manage that, despite the nice lady at the 800
    help number trying over and over to cure. (I suppose I could dig back
    into the programming, but I'm not motivated.)

    Some years ago, I helped extending a library that implements both
    reading (decoding) and sending (generating) IR codes, the primary author
    was quite prolific in extending it to any protocol that he got specificattions and/or samples for. I only wrote a driver part for a not
    yet supported microcontroller, but that was good enough to understand
    some of its workings. Sadly, the project has mostly stalled after 2015.
    The code is still working and small microcontrollers and IR remotes
    don't change that much.

    From your description, I cannot deduce whether a single button press on
    your IR control serves as an "invert the boolean that denotes specific state" (power on/off, for example), or if it is something else. Some universal remotes are just simple and stupid recorders, recording and replaying a bitstream without decoding, perhaps after some signal
    cleanup. Others decode and work from tables.

    From memory, most IR remotes use the NEC protocol, after extracting an abstract code from the bistream, that code essentially is triple (device address, command number, modifier), device denoting a specific tv model,
    for example, command number some arbitrary numbering of the keys on the remote, modifier in this case just a single bit denoting "this is
    comming from a repeating, still pressed keys).

    Usually, the behaviour of a IR remote control is as simple as that.

    So much about the basics. Knowing neither your universal control, nor anything about the remote control in question, I can't even guess what
    is causing that problem. I could tell you what I would do to analyze
    and perhaps solve it, but that won't help you, because you don't have
    the necessary equipment (and knowledge). And even I have stored away
    most of the stuff I need for such work, in order to get space to build
    and maintain our bikes.




    I've seen many such fans, radiant heaters and the like, where the pull
    chain or drawstring had been lost or damaged. But I have rarely
    misplacted an IR remote, simply because there is no point in moving it
    out of the room where the controlled device is located.

    Your rooms must be much less messy than mine!

    Not really. It's just that the remotes sit on top of the devices that
    they control, when not in use. Most of the time, that is.


    ...when we
    did our first tours with Peugeot bicycles bought in 1978, we were young
    and we mostly rode on the flat. A piece of cake, even with only 2 x 5
    and without indexing.

    As I get older and older, I'm trying to get more comfortable with being
    an old guy who avoids mountains. Maybe some guy in a Frank Patterson >>drawing: >>https://www.worthpoint.com/worthopedia/frank-patterson-cycling-artist-500010568

    Well yes, I get the feeling, believe me. I've ridden exactly 4.9 km
    outside, over the last five months. :-/ I'm trying hard not to become comfortable with that. :-) But that's a different story and not one for
    this forum. Just this much: I've never cycled as far and as high on a
    single day as I have done repeatedly since we retired. Becoming old is
    an obstacle, no question. But you can postpone the consequences of
    ageing, at least for a while.

    Exactly. It's easy to forget just how complicated and difficult to
    manufacture "kleenex" components are: Bowden cables, bearing balls,
    tires, chains ...

    There was a fellow in this group that reported on his attempt to do some small-scale tire manufacturing; it was quite a project. Imagine trying
    to build a serviceable derailleur chain with common tools.

    --

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Joy Beeson@21:1/5 to All on Wed Jan 15 01:25:49 2025
    On Tue, 14 Jan 2025 16:04:11 +0700, John B. <slocombjb@gmail.com>
    wrote:


    I don't know whether you know this but when using a stove to heat a
    room put the stove in the opposite side of the room from the chimney.
    Then run the stovepipe from the stove straight up from the stove to
    the ceiling and then along the ceiling, using mounts to separate the
    stove pipe and ceiling, of course, across the room to the chimney.

    That way you get more heat from the same amount of wood.

    When I was a child, my bedroom had a stove-pipe hole in the middle of
    the floor, and a chimney on the east wall. My older sisters said that
    when they were my age, they would put a leg down through the hole when
    our parents had guests, but a piece of sheet metal had been nailed
    over it before I was old enough to take notice.

    --
    Joy Beeson
    joy beeson at centurylink dot net
    http://wlweather.net/PAGEJOY/

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Joy Beeson@21:1/5 to All on Wed Jan 15 01:22:43 2025
    On Mon, 13 Jan 2025 21:13:57 -0800, Jeff Liebermann <jeffl@cruzio.com>
    wrote:

    I'm currently doing something similar to save firewood, which costs
    about $500/cord delivered.


    I found heating with a wood stove quite nice -- it was warmest in the
    living room, where I sat quietly, and coolest in the kitchen where I
    was bustling about.

    I found it amusing that I would take off clothing when I got cold --
    so I could hop onto the rollers and work up some heat. (It helped
    that a large and fluffy cat would lie on the arm that I laid on the
    rollaway bed for balance.)

    Alas, the stove wasn't compatible with our chimney, so we gave it
    away.


    --
    Joy Beeson
    joy beeson at centurylink dot net
    http://wlweather.net/PAGEJOY/

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Ted Heise@21:1/5 to Jeff Liebermann on Wed Jan 15 15:03:44 2025
    On Mon, 13 Jan 2025 21:13:57 -0800,
    Jeff Liebermann <jeffl@cruzio.com> wrote:
    On Mon, 13 Jan 2025 12:58:35 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:
    On 1/13/2025 12:06 PM, Wolfgang Strobl wrote:
    You, as a person, don't need or consume heat. Somewhat
    simplified, you need a certain range of temperatures. In the
    ideal case, you don't need any additional energy, because
    your body already produces heat. A little bit of isolation,
    perhaps provided by that very blanket, might be sufficient.

    And, I suppose, we could do away with all home heating, and
    just wear very heavy clothing all winter. But I don't know of
    anyone doing that.

    I'm currently doing something similar to save firewood, which
    costs about $500/cord delivered.

    Oh my gosh, that's outrageous! A buddy and I cut firewood for a
    living in Flagstaff one summer. We charged $30 for a cord of pine
    (cut, split, delivered and stacked), Oak was $40.

    --
    Ted Heise <theise@panix.com> West Lafayette, IN, USA

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Wolfgang Strobl@21:1/5 to All on Wed Jan 15 17:08:41 2025
    Am Tue, 14 Jan 2025 11:51:17 -0500 schrieb Radey Shouman
    <shouman@comcast.net>:

    Wolfgang Strobl <news51@mystrobl.de> writes:

    Am Mon, 13 Jan 2025 12:27:27 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    On 1/13/2025 9:57 AM, Wolfgang Strobl wrote:
    Am Sun, 12 Jan 2025 21:05:47 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    On 1/12/2025 3:33 PM, Wolfgang Strobl wrote:
    Am Sat, 11 Jan 2025 19:46:50 -0500 schrieb Frank Krygowski
    <frkrygow@sbcglobal.net>:

    To me, a big advantage is the ability to _look_ at a mechanical device >>>>>>> and _see_ what's wrong....

    That, and the fact I can often affect a repair.

    I prefer devices that don't need repair over their lifetime.

    The weakness I see with that is the assumption that "lifetime" is
    defined as "the amount of time it works." if something stops working, >>>>> its lifetime is over! Throw it out!

    That's far too simplistic.

    It depends. For my purposes, I indeed prefer bicycles that may need
    repairs and modifications over their lifetime, for various reasons. I
    change over my lifetime, so do my bicycles. But there are limits. Want >>>> it cheap, longlived, lightweight and functional? Choose any two.



    As I said, I hate the Kleenex ethic - "It's no good any more, just throw >>>>> it away."

    A strawman isn't getting any more pretty, over time. You won't find many >>>> complex products, machines, vehicles or components with an unlimited
    lifetime. Product lifetime has to be planned. There is innovation,
    innovation means change. There are technical limits. So far, I haven't >>>> heard about bicycle tires that tolerate heavy use over a lifetime of 40 >>>> years, as you ask for. To be precise, I don't know of any that I would >>>> like to use or that I would risk using.

    I think my Cannondale touring bike qualifies. Of course I've replaced >>>consumable items like tires, chains, cogs, brake shoes, handlebar tape >>>and occasionally a chainring.

    That way, any bicycle qualifies. There is essentially no part of a
    bicycle that isn't "consumable". If you are lucky, all consumable items
    are consumed at the same time. So you can just buy another bike, call
    it the repaired one and throw out the old one. :-) Given that declaring
    something consumed is a rather arbitrary decision, you have quite a lot
    of slack with that.


    I've made some equipment substitutions
    (saddle, bar-end shifters, "aero" brake levers) but the original >>>equipment is exceptionally durable.

    That is a tautology. Of course the remaining equipment is exeptionally
    durable, otherwise it would have been replaced earlier, for whatever
    reason.


    For one example: The square taper cranks that Tom mocks still work >>>perfectly well. I had to replace the original sealed bottom bracket one >>>time, but there was no confusion about compatibility (and my cranks did >>>not fall off!). The Stronglite roller bearing headset has also lasted >>>decades, with one parts replacement. The SunTour rear derailleur is
    still perfect, although I did cheat a bit. When I powder coated our >>>bikes, I traded my derailleur for my wife's, figuring hers had many
    fewer miles; but both still work just fine. Wheels are not original >>>because I switched from 27" to 700C, but they're 20 years old.

    So why didn't you buy a 40 years old bicycle from somebody who doesn't
    need his bicycle anymore?

    I guess you don't drive a Ford Model T and you don't use an grammophone
    that needs a steel needle for playing shellac records.

    Personally, I am more concerned about how to use a bicycle rather than
    other modes of transport and optimising the bike for that purpose, and I
    am less concerndedabout whether the bike choosen it will last ten,
    twenty or thirty years.

    How long a bicycle lasts depends upon how much it is used and under what
    conditions. A bicycle that lasts more than thirty years is most likely a
    display piece. That some people like you have the time, space and energy
    to maintain a bicycle much longer than its useful life is under normal
    conditions doesn't prove the opposite. That is not an argument against
    repairing, but an argument against repairing, whatever the cost. I'm not
    talking about money only, here. I mostly miss a sense of proportion.

    ...

    As I get older and older, I'm trying to get more comfortable with being >>>an old guy who avoids mountains. Maybe some guy in a Frank Patterson >>>drawing: >>>https://www.worthpoint.com/worthopedia/frank-patterson-cycling-artist-500010568

    Well yes, I get the feeling, believe me. I've ridden exactly 4.9 km
    outside, over the last five months. :-/ I'm trying hard not to become
    comfortable with that. :-) But that's a different story and not one for
    this forum. Just this much: I've never cycled as far and as high on a
    single day as I have done repeatedly since we retired. Becoming old is
    an obstacle, no question. But you can postpone the consequences of
    ageing, at least for a while.

    Exactly. It's easy to forget just how complicated and difficult to >manufacture "kleenex" components are: Bowden cables, bearing balls,
    tires, chains ...

    Indeed, that's part of the point I'm making: many modern products are an aggregation of larger components that are repairable only in theory, but
    not in practice. While most bicycles still are better repairable than
    other products, they still fit this description. Most of the time, you
    just are replacing defective complex components by new ones.



    There was a fellow in this group that reported on his attempt to do some >small-scale tire manufacturing; it was quite a project. Imagine trying
    to build a serviceable derailleur chain with common tools.

    Rohloff, the company that builds and sells the Rohloff SPEEDHUB (<https://www.rohloff.de/en/products/speedhub>) was founded by the
    founder with this machine, for the production of bicycle chains.

    <https://www.rohloff.de/en/company/rohloff-ag/s-l-t-99/>


    --
    Thank you for observing all safety precautions

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Roger Merriman@21:1/5 to Frank Krygowski on Wed Jan 15 17:12:57 2025
    Frank Krygowski <frkrygow@sbcglobal.net> wrote:
    On 1/14/2025 7:12 AM, Roger Merriman wrote:

    The only downside I’ve noted but this is definitely anecdotal than
    statistically significant is of my fellow Gravel club mates which is
    relatively small group two of which have had derailleurs electronic
    failures, both where covered under warranty, but does bring with it a new
    and extra failure point.
    I guess a fair comparison would be to ask how many of your gravel mates
    have had failures of cable actuated mechanical derailleurs.


    None I upgraded my rear mech to GRX to get more range out of the rear mech
    and to get a “clutch” in Shimano speak essentially the rear mech is tighter so keeps the chain away from the chainstay so is a quieter operation on jiggly/rattly bits.

    Roger Merriman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Zen Cycle@21:1/5 to AMuzi on Wed Jan 15 12:34:14 2025
    On 1/14/2025 11:23 AM, AMuzi wrote:
    On 1/14/2025 10:11 AM, cyclintom wrote:
    On Sun Jan 5 05:49:38 2025 zen cycle  wrote:

    I was actually thinking your college years would have predated the
    HP-35, but wasn't sure and was too lazy to look it up.

    I remember the day my father brought home a Bowmar Brain.

    http://www.vintagecalculators.com/html/bowmar_calculators.html

    I was a freshman in high school, would have been 1975.

    That's interesting. You've been telling us that you race. That would
    put you racing at 66 years old. Really competitive.

    More of tommy's "new math". How old would someone who was a freshman in
    1976* be today? anyone?


    You don't know that. I don't either.

    But bicycle races are age-sorted, so racing against other 60
    `70 year olds is as competitive as the 'open' 18 to 40 class.

    +1
    It's all relative. Lining up in the 60+ group today is no less
    competitive that it was lining up in the Master 35+ 25 years ago.



    * I note I typod that, My freshman year was '76/'77
    --
    Add xx to reply

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Zen Cycle@21:1/5 to Frank Krygowski on Wed Jan 15 13:05:19 2025
    On 1/13/2025 11:03 AM, Frank Krygowski wrote:
    On 1/13/2025 8:57 AM, zen cycle wrote:
    On 1/13/2025 2:01 AM, Jeff Liebermann wrote:
    On Mon, 13 Jan 2025 13:20:04 +0700, John B. <slocombjb@gmail.com>
    wrote:

    But long before the days of transistors and "circuit boards" my folks
    had an electric blanket... used it all the rest of their lives. Is the >>>> modern one better?

    Not really better, but certainly safer.  Electric heaters are all 100%
    efficient.  Every watt you shove into the heating wires is turned into
    heat.  There's no way to improve on 100% efficiency.

    Nothing is 100% efficient

    It's certainly true that 100% of the electricity consumed by an electric blanket becomes heat.

    No, that isn't true either. I write "either" because even _if_ it were
    true that electric heaters are 100% efficient (which isn't true), saying
    100% of the electricity consumed by the device become heat is very
    different than saying it's 100% efficient.


    I've always had a mindset for minimizing waste. For most of my life I
    was diligent about shutting off the light switch when leaving a room.
    But now I often don't bother, especially in winter. LEDs draw so much
    less electricity anyway, and in the winter the "waste" just amounts to electric heat.




    --
    Add xx to reply

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Frank Krygowski@21:1/5 to Roger Merriman on Wed Jan 15 13:24:14 2025
    On 1/15/2025 12:12 PM, Roger Merriman wrote:
    Frank Krygowski <frkrygow@sbcglobal.net> wrote:
    On 1/14/2025 7:12 AM, Roger Merriman wrote:

    The only downside I’ve noted but this is definitely anecdotal than
    statistically significant is of my fellow Gravel club mates which is
    relatively small group two of which have had derailleurs electronic
    failures, both where covered under warranty, but does bring with it a new >>> and extra failure point.
    I guess a fair comparison would be to ask how many of your gravel mates
    have had failures of cable actuated mechanical derailleurs.


    None I upgraded my rear mech to GRX to get more range out of the rear mech and to get a “clutch” in Shimano speak essentially the rear mech is tighter
    so keeps the chain away from the chainstay so is a quieter operation on jiggly/rattly bits.
    "None" is good to know. So there were electronic derailleur failures,
    but no mechanical derailleur failures. That gives us a hint about
    relative reliability.

    --
    - Frank Krygowski

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Zen Cycle@21:1/5 to Frank Krygowski on Wed Jan 15 13:28:13 2025
    On 1/15/2025 1:16 PM, Frank Krygowski wrote:
    On 1/15/2025 1:05 PM, Zen Cycle wrote:
    On 1/13/2025 11:03 AM, Frank Krygowski wrote:

    It's certainly true that 100% of the electricity consumed by an
    electric blanket becomes heat.

    No, that isn't true either.

    Please explain. What electrical energy goes elsewhere?

    A very small amount of power is used for the indicator lighting and
    electronic controls.


    I write "either" because even _if_ it were true that electric heaters
    are 100% efficient (which isn't true), saying 100% of the electricity
    consumed by the device become heat is very different than saying it's
    100% efficient.

    What's your definition of "efficiency?" As I said earlier, I think a
    common one used for engineering matters is Desired Output divided by
    Required Input, or something similar.

    Do you have a different one?

    Nope, it's the notion that every watt of power directly goes into
    heating the targeted space that I'm stuck on. There are other losses in
    the cabling and plug interface which - while realized as heat - do not contribute the heating of the targeted space. The heat generated by the
    plug and cord are rather well insulated.

    You may have experienced unplugging an electric appliance and felt the
    plug to be hot. What's happening is that the resistance of the
    plug/socket interface generates heat. In those cases you were able to
    touch the actual metal contacts inside the plug, it's likely it would
    leave a blister. The fact that it doesn't shows the insulation is
    preventing the heat from leaving the system - IOW, not 100% efficiency.





    --
    Add xx to reply

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Frank Krygowski@21:1/5 to Zen Cycle on Wed Jan 15 13:16:01 2025
    On 1/15/2025 1:05 PM, Zen Cycle wrote:
    On 1/13/2025 11:03 AM, Frank Krygowski wrote:

    It's certainly true that 100% of the electricity consumed by an
    electric blanket becomes heat.

    No, that isn't true either.

    Please explain. What electrical energy goes elsewhere?

    I write "either" because even _if_ it were
    true that electric heaters are 100% efficient (which isn't true), saying
    100% of the electricity consumed by the device become heat is very
    different than saying it's 100% efficient.

    What's your definition of "efficiency?" As I said earlier, I think a
    common one used for engineering matters is Desired Output divided by
    Required Input, or something similar.

    Do you have a different one?


    --
    - Frank Krygowski

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Frank Krygowski@21:1/5 to Zen Cycle on Wed Jan 15 13:42:18 2025
    On 1/15/2025 1:28 PM, Zen Cycle wrote:
    On 1/15/2025 1:16 PM, Frank Krygowski wrote:
    On 1/15/2025 1:05 PM, Zen Cycle wrote:
    On 1/13/2025 11:03 AM, Frank Krygowski wrote:

    It's certainly true that 100% of the electricity consumed by an
    electric blanket becomes heat.

    No, that isn't true either.

    Please explain. What electrical energy goes elsewhere?

    A very small amount of power is used for the indicator lighting and electronic controls.


    I write "either" because even _if_ it were true that electric heaters
    are 100% efficient (which isn't true), saying 100% of the electricity
    consumed by the device become heat is very different than saying it's
    100% efficient.

    What's your definition of "efficiency?" As I said earlier, I think a
    common one used for engineering matters is Desired Output divided by
    Required Input, or something similar.

    Do you have a different one?

    Nope, it's the notion that every watt of power directly goes into
    heating the targeted space that I'm stuck on.

    You're moving goalposts. You objected to my statement "It's certainly
    true that 100% of the electricity consumed by an electric blanket
    becomes heat."

    There are other losses in
    the cabling and plug interface which - while realized as heat - do not contribute the heating of the targeted space. The heat generated by the
    plug and cord are rather well insulated.

    But it's still heat, delivered into the room. It's not lost elsewhere.


    You may have experienced unplugging an electric appliance and felt the
    plug to be hot. What's happening is that the resistance of the plug/
    socket interface generates heat. In those cases you were able to touch
    the actual metal contacts inside the plug, it's likely it would leave a blister. The fact that it doesn't shows the insulation is preventing the
    heat from leaving the system - IOW, not 100% efficiency.

    Actually, thermal insulation does not normally prevent heat from leaving
    a system. It merely reduces the rate at which it leaves. That would be
    true of, say, some hot component in a blanket controller. More
    obviously, it's true of the plastic insulation of the heating elements
    within the blanket, and it's true of the fibers of the blanket itself. Nonetheless, all that heat eventually gets delivered. None goes elsewhere.


    --
    - Frank Krygowski

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to All on Wed Jan 15 11:34:04 2025
    On Wed, 15 Jan 2025 12:34:14 -0500, Zen Cycle <funkmaster@hotmail.com>
    wrote:

    On 1/14/2025 11:23 AM, AMuzi wrote:
    On 1/14/2025 10:11 AM, cyclintom wrote:
    On Sun Jan 5 05:49:38 2025 zen cycle wrote:

    I was actually thinking your college years would have predated the
    HP-35, but wasn't sure and was too lazy to look it up.

    I remember the day my father brought home a Bowmar Brain.

    http://www.vintagecalculators.com/html/bowmar_calculators.html

    I was a freshman in high school, would have been 1975.

    That's interesting. You've been telling us that you race. That would
    put you racing at 66 years old. Really competitive.

    More of tommy's "new math". How old would someone who was a freshman in
    1976* be today? anyone?

    A freshman in high skool would typically 14 years old. <https://www.collegevine.com/faq/121389/freshman-age-in-high-school>
    The number of years between high school and today are:
    2025 - 1976 = 49 years.
    The age is therefore:
    49 + 14 = 63 years.

    Please send my prize to the address below.

    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Zen Cycle@21:1/5 to Frank Krygowski on Wed Jan 15 15:42:39 2025
    On 1/15/2025 1:42 PM, Frank Krygowski wrote:
    On 1/15/2025 1:28 PM, Zen Cycle wrote:
    On 1/15/2025 1:16 PM, Frank Krygowski wrote:
    On 1/15/2025 1:05 PM, Zen Cycle wrote:
    On 1/13/2025 11:03 AM, Frank Krygowski wrote:

    It's certainly true that 100% of the electricity consumed by an
    electric blanket becomes heat.

    No, that isn't true either.

    Please explain. What electrical energy goes elsewhere?

    A very small amount of power is used for the indicator lighting and
    electronic controls.


    I write "either" because even _if_ it were true that electric
    heaters are 100% efficient (which isn't true), saying 100% of the
    electricity consumed by the device become heat is very different
    than saying it's 100% efficient.

    What's your definition of "efficiency?" As I said earlier, I think a
    common one used for engineering matters is Desired Output divided by
    Required Input, or something similar.

    Do you have a different one?

    Nope, it's the notion that every watt of power directly goes into
    heating the targeted space that I'm stuck on.

    You're moving goalposts. You objected to my statement "It's certainly
    true that 100% of the electricity consumed by an electric blanket
    becomes heat."

    Nope, I addressed that point very specifically with "A very small amount
    of power is used for the indicator lighting and electronic controls."

    What you misinterpreted as 'moving the goalposts' was me taking issue
    with Jeffs assertion that "electric heaters are all 100% efficient".

    Note that "electric heaters are all 100% efficient" ≠ "It's certainly
    true that 100% of the electricity consumed by an electric blanket
    becomes heat."

    The efficiency of the heater is determined by the energy that is used specifically for generating heat. By that premise, it's logically
    possible that that the heating element in a heating appliance may be
    near 100%, but that some energy will be used for the control portion of
    the system.

    If Jeff had written "Electric heaters are all _nearly_ 100% efficient" I wouldn't have commented.




    There are other losses in the cabling and plug interface which - while
    realized as heat - do not contribute the heating of the targeted
    space. The heat generated by the plug and cord are rather well insulated.

    But it's still heat, delivered into the room. It's not lost elsewhere.

    I differ with "delivered to the room". It's held within the insulation,
    not delivered to the room (see more below).



    You may have experienced unplugging an electric appliance and felt the
    plug to be hot. What's happening is that the resistance of the plug/
    socket interface generates heat. In those cases you were able to touch
    the actual metal contacts inside the plug, it's likely it would leave
    a blister. The fact that it doesn't shows the insulation is preventing
    the heat from leaving the system - IOW, not 100% efficiency.

    Actually, thermal insulation does not normally prevent heat from leaving
    a system. It merely reduces the rate at which it leaves. That would be
    true of, say, some hot component in a blanket controller. More
    obviously, it's true of the plastic insulation of the heating elements
    within the blanket, and it's true of the fibers of the blanket itself. Nonetheless, all that heat eventually gets delivered. None goes elsewhere.

    Hmmm....Is that why the water heater in my basement is still cool to the
    touch 20 years after it was installed? It's been keeping my water at 175 degrees that whole time. By your logic, shouldn't the temperature of the
    outer surface of the tank be 175 Degrees by now? Or at least much warmer
    than the surrounding air?

    The point is that a direct conversion of energy from electrical wattage
    into the system to BTU output won't show 100% efficiency.





    --
    Add xx to reply

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Catrike Ryder@21:1/5 to frkrygow@gXXmail.com on Wed Jan 15 15:44:28 2025
    On Wed, 15 Jan 2025 13:24:14 -0500, Frank Krygowski
    <frkrygow@gXXmail.com> wrote:

    On 1/15/2025 12:12 PM, Roger Merriman wrote:
    Frank Krygowski <frkrygow@sbcglobal.net> wrote:
    On 1/14/2025 7:12 AM, Roger Merriman wrote:

    The only downside Ive noted but this is definitely anecdotal than
    statistically significant is of my fellow Gravel club mates which is
    relatively small group two of which have had derailleurs electronic
    failures, both where covered under warranty, but does bring with it a new >>>> and extra failure point.
    I guess a fair comparison would be to ask how many of your gravel mates
    have had failures of cable actuated mechanical derailleurs.


    None I upgraded my rear mech to GRX to get more range out of the rear mech >> and to get a clutch in Shimano speak essentially the rear mech is tighter >> so keeps the chain away from the chainstay so is a quieter operation on
    jiggly/rattly bits.
    "None" is good to know. So there were electronic derailleur failures,
    but no mechanical derailleur failures. That gives us a hint about
    relative reliability.

    Here's two other ideas.

    1) if you disapprove of electronic shifting, don't buy them.

    2) The fact that you disapprove of them is not going to stop other
    people from buying them.

    --
    C'est bon
    Soloman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Catrike Ryder@21:1/5 to frkrygow@sbcglobal.net on Thu Jan 16 04:21:37 2025
    On Wed, 15 Jan 2025 21:05:22 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/15/2025 3:44 PM, Catrike Ryder wrote:
    On Wed, 15 Jan 2025 13:24:14 -0500, Frank Krygowski
    <frkrygow@gXXmail.com> wrote:

    On 1/15/2025 12:12 PM, Roger Merriman wrote:
    Frank Krygowski <frkrygow@sbcglobal.net> wrote:
    On 1/14/2025 7:12 AM, Roger Merriman wrote:

    The only downside Ive noted but this is definitely anecdotal than >>>>>> statistically significant is of my fellow Gravel club mates which is >>>>>> relatively small group two of which have had derailleurs electronic >>>>>> failures, both where covered under warranty, but does bring with it a new
    and extra failure point.
    I guess a fair comparison would be to ask how many of your gravel mates >>>>> have had failures of cable actuated mechanical derailleurs.


    None I upgraded my rear mech to GRX to get more range out of the rear mech >>>> and to get a clutch in Shimano speak essentially the rear mech is tighter
    so keeps the chain away from the chainstay so is a quieter operation on >>>> jiggly/rattly bits.
    "None" is good to know. So there were electronic derailleur failures,
    but no mechanical derailleur failures. That gives us a hint about
    relative reliability.

    Here's two other ideas.

    1) if you disapprove of electronic shifting, don't buy them.

    2) The fact that you disapprove of them is not going to stop other
    people from buying them.

    IOW, we are not allowed to discuss reliability of bike parts?

    1) Usage by a half dozen or so of Roger's friends doesn't even begin
    to define reliability.

    2) It's already been established that among a half dozen or so of
    Roger's friends, two had electronic shift failures and none had cable
    shift failures, so what more is there to discuss?

    Sheesh. Get over your obsession with me, you miserable old man.

    IOW, am I not allowed to discusss your continuous complaining about
    what other people do?

    It seems to me that it's Krygowski who lives a miserable, unhappy
    existence.

    A good share of his RBT posts involve him whining and complaining
    about something.
    He doesn't like how the bicycling trends are heading;
    he doesn't like the frame designs;
    he doesn't like the newer bicycle components; he doesn't like the
    newer accessories;
    he doesn't like bicycle helmets, flashy biking shirts, or SPD shoes
    and pedals;
    he doesn't like the bicycle infrastructure trends;
    he doesn't like bicycle trails;
    he dislikes dogs and black pickup trucks;
    he doesn't like guns;
    he doesn't like the marketing of anything;
    he complains about people buying lottery tickets;
    he doesn't like how other people select wine or their favorite
    sounding violin;
    he clearly dislikes the current political trends;
    he hates people who ignore his advice;
    he hates that people drive instead riding their bikes to their jobs or
    the grocery store.
    When he's not complaining about something, he trying to get people to
    change their minds and agree with him, which doesn't happen very
    often, so then he whines and complains about that. His brags and lies
    are scattered among his constant complaints and attempts to manipulate
    others.
    His need for imaginary friends is hardly surprising. He has probably
    driven away all the people who know him.


    --
    C'est bon
    Soloman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From zen cycle@21:1/5 to Frank Krygowski on Thu Jan 16 06:50:09 2025
    On 1/13/2025 12:38 PM, Frank Krygowski wrote:
    On 1/13/2025 8:54 AM, zen cycle wrote:
    On 1/12/2025 2:25 PM, Frank Krygowski wrote:
    On 1/12/2025 7:49 AM, zen cycle wrote:
    On 1/11/2025 7:46 PM, Frank Krygowski wrote:

    But individual power outputs from left and right legs? I can't
    imagine much value from that measurement.

    You answered your own question. "You" can't imagine it's useful.
    There are a great many data geeks who find it quite useful.

    Maybe we can take a poll? How many here gather and use data on their
    individual legs' power output?

    I hardly consider this forum to be representative of the cycling
    community at alrge.

    I certainly agree with that! This is a group with no members who were
    not interested in the title "tech".  So if these people are not
    measuring individual leg power, I doubt it's very common.

    That's myopic at best.


    Our bike club is mostly social, not very competitive. I do have friends
    who use "new" equipment (e.g. aero wheels, 13 cogs, road disc brakes)
    and a few who mention their KOM triumphs. Many of them use Garmin or
    similar devices. But IIRC none have ever discussed power meter readings. Maybe some new, younger club members are doing that, but I don't ride
    with those folks.

    Then there's the vast majority of bike riders who will never join a
    club, never look at a bike magazine, never try to ride fast, who just
    ride for fun. No power meters there.


    And in my experience the vast majority of cyclists I ride with use a
    power meter - even those in my age group. Your experience is not the
    norm, Frank, neither is mine.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From zen cycle@21:1/5 to Frank Krygowski on Thu Jan 16 07:14:33 2025
    On 1/15/2025 6:39 PM, Frank Krygowski wrote:
    On 1/15/2025 3:42 PM, Zen Cycle wrote:
    On 1/15/2025 1:42 PM, Frank Krygowski wrote:
    On 1/15/2025 1:28 PM, Zen Cycle wrote:
    On 1/15/2025 1:16 PM, Frank Krygowski wrote:
    On 1/15/2025 1:05 PM, Zen Cycle wrote:
    On 1/13/2025 11:03 AM, Frank Krygowski wrote:

    It's certainly true that 100% of the electricity consumed by an
    electric blanket becomes heat.

    No, that isn't true either.

    Please explain. What electrical energy goes elsewhere?

    A very small amount of power is used for the indicator lighting and
    electronic controls.


    I write "either" because even _if_ it were true that electric
    heaters are 100% efficient (which isn't true), saying 100% of the
    electricity consumed by the device become heat is very different
    than saying it's 100% efficient.

    What's your definition of "efficiency?" As I said earlier, I think
    a common one used for engineering matters is Desired Output divided
    by Required Input, or something similar.

    Do you have a different one?

    Nope, it's the notion that every watt of power directly goes into
    heating the targeted space that I'm stuck on.

    You're moving goalposts. You objected to my statement "It's certainly
    true that 100% of the electricity consumed by an electric blanket
    becomes heat."

    Nope, I addressed that point very specifically with "A very small
    amount of power is used for the indicator lighting and electronic
    controls."

    What you misinterpreted as 'moving the goalposts' was me taking issue
    with Jeffs assertion that "electric heaters are all 100% efficient".

    Note that "electric heaters are all 100% efficient" ≠ "It's certainly
    true that 100% of the electricity consumed by an electric blanket
    becomes heat."

    The efficiency of the heater is determined by the energy that is used
    specifically for generating heat. By that premise, it's logically
    possible that that the heating element in a heating appliance may be
    near 100%, but that some energy will be used for the control portion
    of the system.

    You should look at the energy used for the controls and think about what becomes of it. Do that in microcopic detail.

    If you think it all ends up as heat (IR spectrum) you have a gross misunderstanding of electronics. First off, the indicators dissipate
    energy in the visible light spectrum (this is why LEDs are more
    efficient lighting than incandescent, very little energy is used in the
    IR spectrum). Secondly, think about your premise that it all turns into
    heat - this means no energy is available to do any other form of work.

    Actually, thermal insulation does not normally prevent heat from
    leaving a system. It merely reduces the rate at which it leaves. That
    would be true of, say, some hot component in a blanket controller.
    More obviously, it's true of the plastic insulation of the heating
    elements within the blanket, and it's true of the fibers of the
    blanket itself. Nonetheless, all that heat eventually gets delivered.
    None goes elsewhere.

    Hmmm....Is that why the water heater in my basement is still cool to
    the touch 20 years after it was installed? It's been keeping my water
    at 175 degrees that whole time. By your logic, shouldn't the
    temperature of the outer surface of the tank be 175 Degrees by now? Or
    at least much warmer than the surrounding air?

    I'm happy to discuss this in great detail if you like.

    _IF_ the water heater were enclosed in some big box from which
    absolutely no heat could possibly escape (understand, that's
    impossible), the outside of the water heater would eventually reach the temperature of the water.

    But in the real world, there is heat being lost continually to the atmosphere, etc. in your basement. ("etc" is because a very small amount
    is lost by radiation instead of convection, and is absorbed by solid surroundings.) The primary heat loss, by far, is convection to the
    outside air.

    There is a constant flow of heat energy from within the water heater to
    the air outside. The _rate_ of heat flow depends on the temperature difference (which is analogous to voltage in an electrical circuit) and
    on the amount of insulation (whose "thermal resistance" is analogous to electrical resistance.)

    In practice, the thermal resistance is never infinite. IOW, you can add
    more and more insulation, but you can never reduce heat flow to zero.

    If you were to use absolutely no hot water, and you were to shut off all energy input (gas or electricity or whatever) to your water heater, you
    would eventually find the water at room temperature. Heat would flow
    until the temperature difference across the insulation were zero - very analogous to a capacitor sending _some_ current through even a very high resistance, until the capacitor's voltage was zero.

    OK, that's a fair point, but it seems to me you're agreeing that the
    insulation limits the efficient transfer of heat to the surrounding environment, which was my point. Sure, the outside of the enclosure may _eventually_ reach the temperature of the inner heat element, but if you
    leave a heating element turned on for 24 hours and you haven't
    measurably raised the temperature of the target area one would hardly
    call that an efficient system.


    The point is that a direct conversion of energy from electrical
    wattage into the system to BTU output won't show 100% efficiency.

    We disagree. Again, feel free to explain in detail where you think the
    lost energy would go, if not to heat.



    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Frank Krygowski on Thu Jan 16 11:38:51 2025
    On 1/16/2025 11:18 AM, Frank Krygowski wrote:
    On 1/16/2025 7:14 AM, zen cycle wrote:
    On 1/15/2025 6:39 PM, Frank Krygowski wrote:

    You should look at the energy used for the controls and
    think about what becomes of it. Do that in microcopic
    detail.

    If you think it all ends up as heat (IR spectrum) you have
    a gross misunderstanding of electronics. First off, the
    indicators dissipate energy in the visible light spectrum
    (this is why LEDs are more efficient lighting than
    incandescent, very little energy is used in the IR spectrum).

    OK, a thought experiment: Take an adiabatic container - that
    is, a _perfectly_ insulated box (a physical impossibility,
    but useful for our analysis). Let the box contain whatever
    you like - just air, some solid objects, whatever.

    Cut an LED sized hole in it and insert an LED of your choice
    so it shines into the box. Turn on the LED.

    What happens to the light entering the box? Obviously, you
    don't end up with a box full of light, so it isn't stored;
    it somehow goes away. And what happens to the temperature
    inside the box, and why?

    Answer: The temperature of whatever's inside the box will
    rise. The energy put into _all_ spectra by the LED,
    including the visible light spectrum, ultimately converts to
    heat.

    Secondly, think about your premise that it all turns into
    heat - this means no energy is available to do any other
    form of work.
    In physics or mechanical engineering, work is defined as
    force moving through a distance, or torque moving though an
    angle of rotation. Valid units of measurement are the same
    as the units for energy: foot*pounds, Newton*meters or
    Joules, etc. all of which (interestingly) can be converted
    to BTUs, which are normally units measuring heat.

    And in general, you're right, energy converted to heat is
    not normally available to do work.

    There are exceptions, of course. A device that produces work
    from heat is known as an "engine."

    I bought one of these a while ago, as a geek toy. (I used to
    have a larger one as a demonstrator used in Thermodynamics
    class.)

    https://www.amazon.com/gp/product/B0BL29ZF17/ ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1

    It's a small Stirling Engine driven by the temperature
    difference between the hot bottom plate and the colder top
    plate. It sits on top of a cup of hot water. Heat flows into
    the disk on the bottom and flows out the top one. I've
    gotten about 850 rpm out of mine, more if I place an ice
    cube on the top plate.

    So an engine is a device that converts heat to mechanical
    work. But no engine can covert heat to work with 100%
    efficiency.

    All of this discussion is stuff normally covered under the
    topic "Thermodynamics."


    +1 to all that.

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Wolfgang Strobl@21:1/5 to All on Thu Jan 16 19:12:20 2025
    Am Thu, 16 Jan 2025 06:50:09 -0500 schrieb zen cycle <funkmasterxx@hotmail.com>:

    On 1/13/2025 12:38 PM, Frank Krygowski wrote:

    Then there's the vast majority of bike riders who will never join a
    club, never look at a bike magazine, never try to ride fast, who just
    ride for fun. No power meters there.

    I find the concept of dividing people into “bike riders”, “car drivers” and “train drivers” strange. We have many more “bike riders” than “car
    drivers” in Germany, but most of people have probably used all three
    modes of transportation over the course of the year. Even more strange
    is the idea that one has to be member of a club in order to ride a road
    bike with drop bars or any bike, or to use equipment that helps with
    riding fast or getting fit.

    You won't find a gym or physiotherapy studio around here that doesn't
    have stationary bikes whose main feature is the power meter.

    Unfortunately, I currently have the opportunity to make extensive use of
    the equipment of such a studio, which includes a stationary bicycle
    calld "Milon Q Bike".

    <https://duckduckgo.com/?q=%22Milon+Q+Bike%22&t=ffab&iar=images&iax=images&ia=images>
    The primary target group for stationary bike like those are old people
    or people who have to train in a controlled manner due to illness or
    injury. Personally, I don't use it because I have a much better direct
    drive trainer at home. B



    And in my experience the vast majority of cyclists I ride with use a
    power meter - even those in my age group. Your experience is not the
    norm, Frank, neither is mine.

    I' quite sure that the majority of cyclists that I see riding their
    bikes around my home town don't use a powermeter, simply because these
    are still quite expensive. But that is changing.

    It seems obvious to me that a power meter can be useful for many people
    who cycle to stay fit or to get fit. This is especially true for those
    who do not equate cycling with sport.


    --
    Thank you for observing all safety precautions

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Wolfgang Strobl@21:1/5 to All on Thu Jan 16 19:18:28 2025
    Am Mon, 13 Jan 2025 18:24:35 -0600 schrieb AMuzi <am@yellowjersey.org>:

    On 1/13/2025 3:40 PM, Catrike Ryder wrote:

    [useless fullquote, removed]

    A bicycle frame wears out? Really?
    ...

    Some people have very high standards and toss useful things
    when they are merely scuffed or have a small dent. Other
    people have different standards:

    https://www.yellowjersey.org/mitch.html

    In my case, the frame just broke.


    --
    Thank you for observing all safety precautions

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Radey Shouman@21:1/5 to Frank Krygowski on Thu Jan 16 13:45:57 2025
    Frank Krygowski <frkrygow@sbcglobal.net> writes:

    On 1/15/2025 3:42 PM, Zen Cycle wrote:
    On 1/15/2025 1:42 PM, Frank Krygowski wrote:
    On 1/15/2025 1:28 PM, Zen Cycle wrote:
    On 1/15/2025 1:16 PM, Frank Krygowski wrote:
    On 1/15/2025 1:05 PM, Zen Cycle wrote:
    On 1/13/2025 11:03 AM, Frank Krygowski wrote:

    [ ... ]

    In practice, the thermal resistance is never infinite. IOW, you can
    add more and more insulation, but you can never reduce heat flow to
    zero.

    The funny thing is, if you have insulation with sufficiently low
    resistance around a cylindrical object, eventually adding insulation
    reduces the thermal resistance, because the area available for
    convection keeps increasing. Sometimes this is even practically
    meaningful.

    Just thought I would throw that in, as an excruciating detail.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Radey Shouman@21:1/5 to Frank Krygowski on Thu Jan 16 13:38:25 2025
    Frank Krygowski <frkrygow@gXXmail.com> writes:

    On 1/15/2025 1:28 PM, Zen Cycle wrote:
    On 1/15/2025 1:16 PM, Frank Krygowski wrote:
    On 1/15/2025 1:05 PM, Zen Cycle wrote:
    On 1/13/2025 11:03 AM, Frank Krygowski wrote:

    It's certainly true that 100% of the electricity consumed by an
    electric blanket becomes heat.

    No, that isn't true either.

    Please explain. What electrical energy goes elsewhere?
    A very small amount of power is used for the indicator lighting and
    electronic controls.


    I write "either" because even _if_ it were true that electric
    heaters are 100% efficient (which isn't true), saying 100% of the
    electricity consumed by the device become heat is very different
    than saying it's 100% efficient.

    What's your definition of "efficiency?" As I said earlier, I think
    a common one used for engineering matters is Desired Output divided
    by Required Input, or something similar.

    Do you have a different one?
    Nope, it's the notion that every watt of power directly goes into
    heating the targeted space that I'm stuck on.

    You're moving goalposts. You objected to my statement "It's certainly
    true that 100% of the electricity consumed by an electric blanket
    becomes heat."

    There are other losses in the cabling and plug interface which -
    while realized as heat - do not contribute the heating of the
    targeted space. The heat generated by the plug and cord are rather
    well insulated.

    But it's still heat, delivered into the room. It's not lost elsewhere.

    Not necessarily true. Heat is conducted thermally into the electrical
    wires, which often run inside exterior walls, and can thus be conducted
    to the outdoors without heating a room.

    But these are quibbles. The definition of efficiency depends on the
    purpose of the device, and the theoretical model used to compute the
    minimum energy (or whatever) required to achieve that purpose.

    The purpose of an electric blanket is *not* to heat a room, it is to
    make an individual human being more comfortable *without* heating the
    room. Grandpa can feel warm without requiring the thermostat at 90F.

    Note that the "blanket" part of an electric blanket is important.
    Without insulation most of the heat produced electrically will be wasted
    by heating the cold room, since the room is cooler than the person
    warmed. Most people that actually use electric blankets put more
    blankets over them for this reason.

    What is the theoretical minimum heat required to make an individual feel
    warm? Every living human being produces metabolic heat constantly,
    which must be lost to the environment to prevent overheating. A 2000
    Kcal/day diet implies an average power output of 97W, almost all of
    which is heat for a typical sedentary person.

    The minimum heat required is the heat inevitably lost to the environment
    less metabolic heat. For a typical electric blanket application I'm
    guessing the heat lost is mostly due to respiration (sensible heat and
    latent heat of the moisture added to exhaled air).

    You're welcome to compute that; it should be easy once you estimate the volumetric rate of respiration and assume a room temperature and
    humidity. I'm pretty sure the resulting efficiency of an electric
    blanket is way below 100%.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Zen Cycle@21:1/5 to Frank Krygowski on Thu Jan 16 13:58:27 2025
    On 1/16/2025 12:18 PM, Frank Krygowski wrote:
    On 1/16/2025 7:14 AM, zen cycle wrote:
    On 1/15/2025 6:39 PM, Frank Krygowski wrote:

    You should look at the energy used for the controls and think about
    what becomes of it. Do that in microcopic detail.

    If you think it all ends up as heat (IR spectrum) you have a gross
    misunderstanding of electronics. First off, the indicators dissipate
    energy in the visible light spectrum (this is why LEDs are more
    efficient lighting than incandescent, very little energy is used in
    the IR spectrum).

    OK, a thought experiment: Take an adiabatic container - that is, a _perfectly_ insulated box (a physical impossibility, but useful for our analysis). Let the box contain whatever you like - just air, some solid objects, whatever.

    Cut an LED sized hole in it and insert an LED of your choice so it
    shines into the box. Turn on the LED.

    What happens to the light entering the box? Obviously, you don't end up
    with a box full of light, so it isn't stored; it somehow goes away.

    lol...no, it doesn't 'go away'. For all intents and purposes it suffers
    the safe fate as Schrödinger's cat.

    And
    what happens to the temperature inside the box, and why?

    It will rise somewhat due to the residual IR energy from the system.
    Visible spectrum LEDs do emit some IR, just not nearly as much as
    incandescents (given the same lux)


    Answer: The temperature of whatever's inside the box will rise. The
    energy put into _all_ spectra by the LED, including the visible light spectrum, ultimately converts to heat.

    So _all_ forms of energy convert to heat? You should write a paper on that.


    Secondly, think about your premise that it all turns into heat - this
    means no energy is available to do any other form of work.
    In physics or mechanical engineering, work is defined as force moving
    through a distance, or torque moving though an angle of rotation. Valid
    units of measurement are the same as the units for energy: foot*pounds, Newton*meters or Joules, etc. all of which (interestingly) can be
    converted to BTUs, which are normally units measuring heat.

    And in general, you're right, energy converted to heat is not normally available to do work.

    I used the term 'work' more generically, in this case running a program, setting bits in memory, etc. Yes, residual heat from the process, but
    energy is used to perform whatever task, Heat is the result of losses in
    the system (thermal junctions from die bonds, for example)


    There are exceptions, of course. A device that produces work from heat
    is known as an "engine."

    I bought one of these a while ago, as a geek toy. (I used to have a
    larger one as a demonstrator used in Thermodynamics class.)

    https://www.amazon.com/gp/product/B0BL29ZF17/ ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1

    It's a small Stirling Engine driven by the temperature difference
    between the hot bottom plate and the colder top plate. It sits on top of
    a cup of hot water. Heat flows into the disk on the bottom and flows out
    the top one. I've gotten about 850 rpm out of mine, more if I place an
    ice cube on the top plate.

    So an engine is a device that converts heat to mechanical work. But no
    engine can covert heat to work with 100% efficiency.

    All of this discussion is stuff normally covered under the topic "Thermodynamics."



    --
    Add xx to reply

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Wolfgang Strobl on Thu Jan 16 13:07:24 2025
    On 1/16/2025 12:18 PM, Wolfgang Strobl wrote:
    Am Mon, 13 Jan 2025 18:24:35 -0600 schrieb AMuzi <am@yellowjersey.org>:

    On 1/13/2025 3:40 PM, Catrike Ryder wrote:

    [useless fullquote, removed]

    A bicycle frame wears out? Really?
    ...

    Some people have very high standards and toss useful things
    when they are merely scuffed or have a small dent. Other
    people have different standards:

    https://www.yellowjersey.org/mitch.html

    In my case, the frame just broke.




    A (one) frame break?
    That's not a crisis, that's a repair!

    https://www.yellowjersey.org/paint.html

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Radey Shouman on Thu Jan 16 13:13:15 2025
    On 1/16/2025 12:38 PM, Radey Shouman wrote:
    Frank Krygowski <frkrygow@gXXmail.com> writes:

    On 1/15/2025 1:28 PM, Zen Cycle wrote:
    On 1/15/2025 1:16 PM, Frank Krygowski wrote:
    On 1/15/2025 1:05 PM, Zen Cycle wrote:
    On 1/13/2025 11:03 AM, Frank Krygowski wrote:

    It's certainly true that 100% of the electricity consumed by an
    electric blanket becomes heat.

    No, that isn't true either.

    Please explain. What electrical energy goes elsewhere?
    A very small amount of power is used for the indicator lighting and
    electronic controls.


    I write "either" because even _if_ it were true that electric
    heaters are 100% efficient (which isn't true), saying 100% of the
    electricity consumed by the device become heat is very different
    than saying it's 100% efficient.

    What's your definition of "efficiency?" As I said earlier, I think
    a common one used for engineering matters is Desired Output divided
    by Required Input, or something similar.

    Do you have a different one?
    Nope, it's the notion that every watt of power directly goes into
    heating the targeted space that I'm stuck on.

    You're moving goalposts. You objected to my statement "It's certainly
    true that 100% of the electricity consumed by an electric blanket
    becomes heat."

    There are other losses in the cabling and plug interface which -
    while realized as heat - do not contribute the heating of the
    targeted space. The heat generated by the plug and cord are rather
    well insulated.

    But it's still heat, delivered into the room. It's not lost elsewhere.

    Not necessarily true. Heat is conducted thermally into the electrical
    wires, which often run inside exterior walls, and can thus be conducted
    to the outdoors without heating a room.

    But these are quibbles. The definition of efficiency depends on the
    purpose of the device, and the theoretical model used to compute the
    minimum energy (or whatever) required to achieve that purpose.

    The purpose of an electric blanket is *not* to heat a room, it is to
    make an individual human being more comfortable *without* heating the
    room. Grandpa can feel warm without requiring the thermostat at 90F.

    Note that the "blanket" part of an electric blanket is important.
    Without insulation most of the heat produced electrically will be wasted
    by heating the cold room, since the room is cooler than the person
    warmed. Most people that actually use electric blankets put more
    blankets over them for this reason.

    What is the theoretical minimum heat required to make an individual feel warm? Every living human being produces metabolic heat constantly,
    which must be lost to the environment to prevent overheating. A 2000 Kcal/day diet implies an average power output of 97W, almost all of
    which is heat for a typical sedentary person.

    The minimum heat required is the heat inevitably lost to the environment
    less metabolic heat. For a typical electric blanket application I'm
    guessing the heat lost is mostly due to respiration (sensible heat and
    latent heat of the moisture added to exhaled air).

    You're welcome to compute that; it should be easy once you estimate the volumetric rate of respiration and assume a room temperature and
    humidity. I'm pretty sure the resulting efficiency of an electric
    blanket is way below 100%.

    Datuayins are on different channels.

    Mr Krygowski is right regarding physics and the laws which
    yet obtain.

    If you want to discuss efficiency in terms of product
    marketing (New! Heats gramps, not the room!) that's a
    different area.

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to AMuzi on Thu Jan 16 13:39:16 2025
    On 1/16/2025 1:13 PM, AMuzi wrote:
    On 1/16/2025 12:38 PM, Radey Shouman wrote:
    Frank Krygowski <frkrygow@gXXmail.com> writes:

    On 1/15/2025 1:28 PM, Zen Cycle wrote:
    On 1/15/2025 1:16 PM, Frank Krygowski wrote:
    On 1/15/2025 1:05 PM, Zen Cycle wrote:
    On 1/13/2025 11:03 AM, Frank Krygowski wrote:

    It's certainly true that 100% of the electricity
    consumed by an
    electric blanket becomes heat.

    No, that isn't true either.

    Please explain. What electrical energy goes elsewhere?
    A very small amount of power is used for the indicator
    lighting and
    electronic controls.


    I write "either" because even _if_ it were true that
    electric
    heaters are 100% efficient (which isn't true), saying
    100% of the
    electricity consumed by the device become heat is very
    different
    than saying it's 100% efficient.

    What's your definition of "efficiency?" As I said
    earlier, I think
    a common one used for engineering matters is Desired
    Output divided
    by Required Input, or something similar.

    Do you have a different one?
    Nope, it's the notion that every watt of power directly
    goes into
    heating the targeted space that I'm stuck on.

    You're moving goalposts. You objected to my statement
    "It's certainly
    true that 100% of the electricity consumed by an electric
    blanket
    becomes heat."

    There are other losses in the cabling and plug interface
    which -
    while realized as heat - do not contribute the heating
    of the
    targeted space. The heat generated by the plug and cord
    are rather
    well insulated.

    But it's still heat, delivered into the room. It's not
    lost elsewhere.

    Not necessarily true.  Heat is conducted thermally into
    the electrical
    wires, which often run inside exterior walls, and can thus
    be conducted
    to the outdoors without heating a room.

    But these are quibbles.  The definition of efficiency
    depends on the
    purpose of the device, and the theoretical model used to
    compute the
    minimum energy (or whatever) required to achieve that
    purpose.

    The purpose of an electric blanket is *not* to heat a
    room, it is to
    make an individual human being more comfortable *without*
    heating the
    room.  Grandpa can feel warm without requiring the
    thermostat at 90F.

    Note that the "blanket" part of an electric blanket is
    important.
    Without insulation most of the heat produced electrically
    will be wasted
    by heating the cold room, since the room is cooler than
    the person
    warmed.  Most people that actually use electric blankets
    put more
    blankets over them for this reason.

    What is the theoretical minimum heat required to make an
    individual feel
    warm?  Every living human being produces metabolic heat
    constantly,
    which must be lost to the environment to prevent
    overheating.  A 2000
    Kcal/day diet implies an average power output of 97W,
    almost all of
    which is heat for a typical sedentary person.

    The minimum heat required is the heat inevitably lost to
    the environment
    less metabolic heat.  For a typical electric blanket
    application I'm
    guessing the heat lost is mostly due to respiration
    (sensible heat and
    latent heat of the moisture added to exhaled air).

    You're welcome to compute that; it should be easy once you
    estimate the
    volumetric rate of respiration and assume a room
    temperature and
    humidity.  I'm pretty sure the resulting efficiency of an
    electric
    blanket is way below 100%.

    Datuayins are on different channels.

    Mr Krygowski is right regarding physics and the laws which
    yet obtain.

    If you want to discuss efficiency in terms of product
    marketing (New! Heats gramps, not the room!) that's a
    different area.


    Apologies to Mr Shouman for 'the two of you'. The back and
    forth was between Mr Krygowski and Mr Zen.

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Catrike Ryder@21:1/5 to frkrygow@sbcglobal.net on Thu Jan 16 15:40:39 2025
    On Thu, 16 Jan 2025 11:49:41 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/16/2025 4:21 AM, Catrike Ryder wrote:
    On Wed, 15 Jan 2025 21:05:22 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    IOW, we are not allowed to discuss reliability of bike parts?

    1) Usage by a half dozen or so of Roger's friends doesn't even begin
    to define reliability.
    Gosh, does that mean we should look at actual large population _data_????

    If you think it's worth the effort, go right ahead. As for me, I don't
    give a shit.

    Oh, and who is "we?" You and "Fred?"

    Why, that can't be right! It goes against half the arguments you've ever
    made here! :-)

    --
    C'est bon
    Soloman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Catrike Ryder@21:1/5 to AMuzi on Thu Jan 16 16:33:20 2025
    On Thu, 16 Jan 2025 13:07:24 -0600, AMuzi <am@yellowjersey.org> wrote:

    On 1/16/2025 12:18 PM, Wolfgang Strobl wrote:
    Am Mon, 13 Jan 2025 18:24:35 -0600 schrieb AMuzi <am@yellowjersey.org>:

    On 1/13/2025 3:40 PM, Catrike Ryder wrote:

    [useless fullquote, removed]

    A bicycle frame wears out? Really?
    ...

    Some people have very high standards and toss useful things
    when they are merely scuffed or have a small dent. Other
    people have different standards:

    https://www.yellowjersey.org/mitch.html

    In my case, the frame just broke.




    A (one) frame break?
    That's not a crisis, that's a repair!

    https://www.yellowjersey.org/paint.html

    My original Catrike frame developed hairline cracks where I'd applied
    some clamps. Catrike wanted the frame back so they could beef up that
    part of the frame. They sent me a new frame and even offered to have
    it all put back together. I opted to do it myself.

    --
    C'est bon
    Soloman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Frank Krygowski on Thu Jan 16 15:39:23 2025
    On 1/16/2025 3:25 PM, Frank Krygowski wrote:
    On 1/16/2025 1:58 PM, Zen Cycle wrote:
    On 1/16/2025 12:18 PM, Frank Krygowski wrote:
    On 1/16/2025 7:14 AM, zen cycle wrote:
    On 1/15/2025 6:39 PM, Frank Krygowski wrote:

    You should look at the energy used for the controls and
    think about what becomes of it. Do that in microcopic
    detail.

    If you think it all ends up as heat (IR spectrum) you
    have a gross misunderstanding of electronics. First off,
    the indicators dissipate energy in the visible light
    spectrum (this is why LEDs are more efficient lighting
    than incandescent, very little energy is used in the IR
    spectrum).

    OK, a thought experiment: Take an adiabatic container -
    that is, a _perfectly_ insulated box (a physical
    impossibility, but useful for our analysis). Let the box
    contain whatever you like - just air, some solid objects,
    whatever.

    Cut an LED sized hole in it and insert an LED of your
    choice so it shines into the box. Turn on the LED.

    What happens to the light entering the box? Obviously,
    you don't end up with a box full of light, so it isn't
    stored; it somehow goes away.

    lol...no, it doesn't 'go away'. For all intents and
    purposes it suffers the safe fate as Schrödinger's cat.

    No cats necessary. If you opened such a box, would you
    really expect to see a bright flash of the released light?
    If so, you'd be disappointed.


    And what happens to the temperature inside the box, and why?

    It will rise somewhat due to the residual IR energy from
    the system. Visible spectrum LEDs do emit some IR, just
    not nearly as much as incandescents (given the same lux)

    Even if only visible spectrum light is emitted, the
    temperature will rise. The visible light energy will be
    converted to heat.

    So _all_ forms of energy convert to heat? You should write
    a paper on that.

    Those papers were written long, long ago. See https:// www.physlink.com/education/askexperts/ae261.cfm for some hints.

    "That's a famous question that people thought a lot about in
    the nineteenth century. It goes under the name of the 'Heat
    Death of the Universe.' In short, once all of the energy in
    the universe is converted to heat then the universe will be
    in equilibrium -- everything will be of the same temperature
    and entropy will remain constant forever."


    +1 again. That's how our world actually is.

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Zen Cycle@21:1/5 to Frank Krygowski on Thu Jan 16 16:54:16 2025
    On 1/16/2025 4:25 PM, Frank Krygowski wrote:
    On 1/16/2025 1:58 PM, Zen Cycle wrote:
    On 1/16/2025 12:18 PM, Frank Krygowski wrote:
    On 1/16/2025 7:14 AM, zen cycle wrote:
    On 1/15/2025 6:39 PM, Frank Krygowski wrote:

    You should look at the energy used for the controls and think about
    what becomes of it. Do that in microcopic detail.

    If you think it all ends up as heat (IR spectrum) you have a gross
    misunderstanding of electronics. First off, the indicators dissipate
    energy in the visible light spectrum (this is why LEDs are more
    efficient lighting than incandescent, very little energy is used in
    the IR spectrum).

    OK, a thought experiment: Take an adiabatic container - that is, a
    _perfectly_ insulated box (a physical impossibility, but useful for
    our analysis). Let the box contain whatever you like - just air, some
    solid objects, whatever.

    Cut an LED sized hole in it and insert an LED of your choice so it
    shines into the box. Turn on the LED.

    What happens to the light entering the box? Obviously, you don't end
    up with a box full of light, so it isn't stored; it somehow goes away.

    lol...no, it doesn't 'go away'. For all intents and purposes it
    suffers the safe fate as Schrödinger's cat.

    No cats necessary. If you opened such a box, would you really expect to
    see a bright flash of the released light? If so, you'd be disappointed.


    And what happens to the temperature inside the box, and why?

    It will rise somewhat due to the residual IR energy from the system.
    Visible spectrum LEDs do emit some IR, just not nearly as much as
    incandescents (given the same lux)

    Even if only visible spectrum light is emitted, the temperature will
    rise. The visible light energy will be converted to heat.

    So _all_ forms of energy convert to heat? You should write a paper on
    that.

    Those papers were written long, long ago. See https://www.physlink.com/ education/askexperts/ae261.cfm for some hints.

    "That's a famous question that people thought a lot about in the
    nineteenth century. It goes under the name of the 'Heat Death of the Universe.' In short, once all of the energy in the universe is converted
    to heat then the universe will be in equilibrium -- everything will be
    of the same temperature and entropy will remain constant forever."



    OK, we're having a semantic quibble. I was working strictly within the
    IR spectrum as sensible heat, you're wrapping in latent heat.

    --
    Add xx to reply

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Frank Krygowski on Thu Jan 16 15:58:57 2025
    On 1/16/2025 3:35 PM, Frank Krygowski wrote:
    On 1/16/2025 1:38 PM, Radey Shouman wrote:
    Frank Krygowski <frkrygow@gXXmail.com> writes:

    On 1/15/2025 1:28 PM, Zen Cycle wrote:
    On 1/15/2025 1:16 PM, Frank Krygowski wrote:
    On 1/15/2025 1:05 PM, Zen Cycle wrote:
    On 1/13/2025 11:03 AM, Frank Krygowski wrote:

    It's certainly true that 100% of the electricity
    consumed by an
    electric blanket becomes heat.

    No, that isn't true either.

    Please explain. What electrical energy goes elsewhere?
    A very small amount of power is used for the indicator
    lighting and
    electronic controls.


    I write "either" because even _if_ it were true that
    electric
    heaters are 100% efficient (which isn't true), saying
    100% of the
    electricity consumed by the device become heat is very
    different
    than saying it's 100% efficient.

    What's your definition of "efficiency?" As I said
    earlier, I think
    a common one used for engineering matters is Desired
    Output divided
    by Required Input, or something similar.

    Do you have a different one?
    Nope, it's the notion that every watt of power directly
    goes into
    heating the targeted space that I'm stuck on.

    You're moving goalposts. You objected to my statement
    "It's certainly
    true that 100% of the electricity consumed by an electric
    blanket
    becomes heat."

    There are other losses in the cabling and plug interface
    which -
    while realized as heat - do not contribute the heating
    of the
    targeted space. The heat generated by the plug and cord
    are rather
    well insulated.

    But it's still heat, delivered into the room. It's not
    lost elsewhere.

    Not necessarily true.  Heat is conducted thermally into
    the electrical
    wires, which often run inside exterior walls, and can thus
    be conducted
    to the outdoors without heating a room.

    But these are quibbles.  The definition of efficiency
    depends on the
    purpose of the device, and the theoretical model used to
    compute the
    minimum energy (or whatever) required to achieve that
    purpose.

    The purpose of an electric blanket is *not* to heat a
    room, it is to
    make an individual human being more comfortable *without*
    heating the
    room.

    I guess it's possible to define the Desired Output more and
    more narrowly, down to "The heat delivered to the parts of
    the body that have nerve endings that detect temperature."
    IOW, if Grandpa's hair and toenails get warmer, that's
    wasted heat. But I think few people want to go to that extreme.

    Slightly more reasonable would be to demand wrapping the
    electric blanket around Grandpa, like a sleeping bag, then
    wrapping that with a perfectly adiabatic blanket. All the
    heat would eventually go into Grandpa.


    +1
    That's marketing not physics.

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Radey Shouman@21:1/5 to Frank Krygowski on Thu Jan 16 17:47:23 2025
    Frank Krygowski <frkrygow@sbcglobal.net> writes:

    On 1/16/2025 1:38 PM, Radey Shouman wrote:
    Frank Krygowski <frkrygow@gXXmail.com> writes:

    On 1/15/2025 1:28 PM, Zen Cycle wrote:
    On 1/15/2025 1:16 PM, Frank Krygowski wrote:
    On 1/15/2025 1:05 PM, Zen Cycle wrote:
    On 1/13/2025 11:03 AM, Frank Krygowski wrote:

    It's certainly true that 100% of the electricity consumed by an
    electric blanket becomes heat.

    No, that isn't true either.

    Please explain. What electrical energy goes elsewhere?
    A very small amount of power is used for the indicator lighting and
    electronic controls.


    I write "either" because even _if_ it were true that electric
    heaters are 100% efficient (which isn't true), saying 100% of the
    electricity consumed by the device become heat is very different
    than saying it's 100% efficient.

    What's your definition of "efficiency?" As I said earlier, I think
    a common one used for engineering matters is Desired Output divided
    by Required Input, or something similar.

    Do you have a different one?
    Nope, it's the notion that every watt of power directly goes into
    heating the targeted space that I'm stuck on.

    You're moving goalposts. You objected to my statement "It's certainly
    true that 100% of the electricity consumed by an electric blanket
    becomes heat."

    There are other losses in the cabling and plug interface which -
    while realized as heat - do not contribute the heating of the
    targeted space. The heat generated by the plug and cord are rather
    well insulated.

    But it's still heat, delivered into the room. It's not lost elsewhere.
    Not necessarily true. Heat is conducted thermally into the
    electrical
    wires, which often run inside exterior walls, and can thus be conducted
    to the outdoors without heating a room.
    But these are quibbles. The definition of efficiency depends on the
    purpose of the device, and the theoretical model used to compute the
    minimum energy (or whatever) required to achieve that purpose.
    The purpose of an electric blanket is *not* to heat a room, it is to
    make an individual human being more comfortable *without* heating the
    room.

    I guess it's possible to define the Desired Output more and more
    narrowly, down to "The heat delivered to the parts of the body that
    have nerve endings that detect temperature." IOW, if Grandpa's hair
    and toenails get warmer, that's wasted heat. But I think few people
    want to go to that extreme.

    Slightly more reasonable would be to demand wrapping the electric
    blanket around Grandpa, like a sleeping bag, then wrapping that with a perfectly adiabatic blanket. All the heat would eventually go into
    Grandpa.

    The only thing I demand is a meaningful definition of efficiency for an electric blanket. I proposed one, and I haven't seen any others.

    Your "perfectly adiabatic blanket" does not address respiration, unless
    it has a perfect heat exchanger between air in and air out, which is
    water saturated. If it worked, then Grandpa would die of overheating in
    short order, without an electric blanket, cold blooded though he might
    be.

    --

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Catrike Ryder@21:1/5 to frkrygow@sbcglobal.net on Thu Jan 16 17:49:01 2025
    On Thu, 16 Jan 2025 16:57:27 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/16/2025 3:40 PM, Catrike Ryder wrote:
    On Thu, 16 Jan 2025 11:49:41 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/16/2025 4:21 AM, Catrike Ryder wrote:
    On Wed, 15 Jan 2025 21:05:22 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    IOW, we are not allowed to discuss reliability of bike parts?

    1) Usage by a half dozen or so of Roger's friends doesn't even begin
    to define reliability.
    Gosh, does that mean we should look at actual large population _data_???? >>
    If you think it's worth the effort, go right ahead. As for me, I don't
    give a shit.

    Oh, and who is "we?" You and "Fred?"

    Why, that can't be right! It goes against half the arguments you've ever >>> made here! :-)


    This guy needs counseling, to get rid of his obsession with me. What a >miserable old man!

    Cut it our, Krygowski. You're the bitterest old coot I've ever come
    across. You whine and complain about everything you come across even
    when it has no tangible effect on you. You haven't expressed a moment
    of contentment since I first began in this group.

    Do you sit up all evening looking for something to be unhappy about?
    Do your wife and family share your discontentment or do you grouch at
    them, too?

    --
    C'est bon
    Soloman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Radey Shouman@21:1/5 to Zen Cycle on Thu Jan 16 17:56:25 2025
    Zen Cycle <funkmaster@hotmail.com> writes:

    On 1/16/2025 12:18 PM, Frank Krygowski wrote:
    On 1/16/2025 7:14 AM, zen cycle wrote:
    On 1/15/2025 6:39 PM, Frank Krygowski wrote:

    You should look at the energy used for the controls and think
    about what becomes of it. Do that in microcopic detail.

    If you think it all ends up as heat (IR spectrum) you have a gross
    misunderstanding of electronics. First off, the indicators
    dissipate energy in the visible light spectrum (this is why LEDs
    are more efficient lighting than incandescent, very little energy
    is used in the IR spectrum).
    OK, a thought experiment: Take an adiabatic container - that is, a
    _perfectly_ insulated box (a physical impossibility, but useful for
    our analysis). Let the box contain whatever you like - just air,
    some solid objects, whatever.
    Cut an LED sized hole in it and insert an LED of your choice so it
    shines into the box. Turn on the LED.
    What happens to the light entering the box? Obviously, you don't end
    up with a box full of light, so it isn't stored; it somehow goes
    away.

    lol...no, it doesn't 'go away'. For all intents and purposes it
    suffers the safe fate as Schrödinger's cat.

    And what happens to the temperature inside the box, and why?

    It will rise somewhat due to the residual IR energy from the
    system. Visible spectrum LEDs do emit some IR, just not nearly as much
    as incandescents (given the same lux)

    Answer: The temperature of whatever's inside the box will rise. The
    energy put into _all_ spectra by the LED, including the visible
    light spectrum, ultimately converts to heat.

    So _all_ forms of energy convert to heat? You should write a paper on that.


    Secondly, think about your premise that it all turns into heat -
    this means no energy is available to do any other form of work.
    In physics or mechanical engineering, work is defined as force
    moving through a distance, or torque moving though an angle of
    rotation. Valid units of measurement are the same as the units for
    energy: foot*pounds, Newton*meters or Joules, etc. all of which
    (interestingly) can be converted to BTUs, which are normally units
    measuring heat.
    And in general, you're right, energy converted to heat is not
    normally available to do work.

    I used the term 'work' more generically, in this case running a
    program, setting bits in memory, etc. Yes, residual heat from the
    process, but energy is used to perform whatever task, Heat is the
    result of losses in the system (thermal junctions from die bonds, for example)

    I don't believe that is correct. All the energy used to run a program
    does eventually get rejected as heat. I suppose either 1 or 0 bits must
    have a slightly higher potential energy, but the net number of each is
    not likely to change much, and the energy difference must be small.

    There is a field of study called "thermodynamic computing", which
    studies the minimum energy that must be dissipated for various logic operations. Turns out that reversible computations, ie those that do
    not irretrievably lose information, can be more thermodynamically
    efficient than those that do lose information. Eventually this sort of thinking is hoped to enable more economical computation.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to All on Thu Jan 16 21:08:46 2025
    On Wed, 15 Jan 2025 13:48:14 +0700, John B. <slocombjb@gmail.com>
    wrote:

    On Tue, 14 Jan 2025 07:40:45 -0800, Jeff Liebermann <jeffl@cruzio.com>
    wrote:

    On Tue, 14 Jan 2025 16:04:11 +0700, John B. <slocombjb@gmail.com>
    wrote:

    I don't know whether you know this but when using a stove to heat a
    room put the stove in the opposite side of the room from the chimney. >>>Then run the stovepipe from the stove straight up from the stove to
    the ceiling and then along the ceiling, using mounts to separate the >>>stove pipe and ceiling, of course, across the room to the chimney.

    That way you get more heat from the same amount of wood.

    Bad idea. When I bought the house in 1974(?), that's roughly what the >>previous owner had done. I couldn't find a photo, but it was a rather >>large "pot belly" stove that leaked air from every seam. It was
    located at one end of the living room with about a 10 ft horizontal
    flue pipe to where it connected to the a 5 ft vertical insulated
    external pipe (Metalbestos) that acted as a chimney pipe. It didn't
    work for a variety of reasons. The big one was that for the "draw" or >>suction to be able to move sufficient hot air up the chimney pipe, the >>inside of the chimney pipe had to be hot. In order to heat the pipe,
    the hot air from the stove has to rise, which is rather awkward with a
    10 ft horizontal uninsulated pipe with nowhere for the hot air to
    rise. Getting a fire started was possible, but difficult and VERY
    smoky. Cleaning the ash out of the horizontal section was difficult >>because the previous owner had neglected to install a "T" where the
    pipe changed from horizontal to vertical. Even if there was a "T", it >>wouldn't have worked because rising hot air does not like making sharp >>turns.

    In other words, it didn't work and horizontal flue pipe are a really
    bad idea. I had to remodel the living room, move the stairs going
    through the floor, replace the pot belly with an airtight stove, and
    extend the external chimney pipe to 12 ft, in order to fix the
    problems. Although I don't completely understand how it works, I've
    found that the "efficiency" of the system is controlled by how much
    time the hot gasses remain in the stove and whether the wood is >>sufficiently dry (under 10% moisture).

    Your description doesn't sound like the systems built in and used in
    about every building in my home town.

    The US east and west coast climates are very different.

    Often built from red brick these
    chimney reached from the cellar to above the roof top.

    In 1989, we had the Loma Prieta Earthquake (Magnitude 7.1). In my neighborhood, Most of the brick chimneys cracked at the mortar lines,
    broke off at the roof line, or fell onto the house destroying the
    roof. The only chimneys that survived were those that were in some
    way reinforced with rebar or had a steel pipe lining. At the time, I
    had two chimneys. One was 8 inch in triple wall DuraVent for the
    downstairs rooms. The upstairs chimney was 6 inch Metalbestos double
    wall. The upstairs Metalbestos chimney had to be straightened. The
    longer downstairs Matalbestos collapsed and ended up on the ground in
    pieces. 8 inch pipe was too large for the downstairs stove, so I sold
    the pieces to a neighbor. I eventually removed most of the 8 inch
    pipe and decided not to replace it because it was workshop and storage
    which didn't need heating. When you live in earthquake country,
    construction practices tend to be rather different.

    Then had a
    small opening in the bottom often with a loose fitting metal door to
    allow cleaning anything they might fall down the chimney and to allow
    the free flow of air up through the chimney. Thus the hotter, lighter, >"smoke" flowed up the stack.

    Yes. My parents house in Smog Angeles was like that. However, the
    brick fireplace was only used during the winter holidays. The rest of
    the year, a natural gas furnace was sufficient.

    As for an air tight stove that is a bit puzzling. How do you have a
    fire in an air tight enclose?

    Air tight is a rather bad name. What it means is that the air intake
    to the stove could be controlled with the air intake almost completely
    closed. That allows me to slowly burn large logs. Having air coming
    into the stove from cracks in the firebox and around the door causes
    problem with control, where the oscillates between almost going out
    and flaring up. With no air leaks, I can set it and forget it.

    Incidentally, shutting down the stove by strangling the air intake
    causes other problems. The fire produces large amounts of CO (carbon
    monoxide) and creosote (flammable tar) if there's not enough air. It's
    also the way one makes charcoal and char cloth (for starting a camp
    fire with a flint and steel).
    <https://en.wikipedia.org/wiki/Char_cloth>

    EPA approved wood burning stoves have two small air intakes that
    cannot be blocked or closed down that effectively prevent producing
    carbon monoxide and creosote. My current stove is not up to date on
    EPA approvals, which generally requires a catalytic converter.



    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to Soloman@old.bikers.org on Thu Jan 16 21:33:25 2025
    On Mon, 13 Jan 2025 14:33:13 -0500, Catrike Ryder
    <Soloman@old.bikers.org> wrote:

    On Mon, 13 Jan 2025 11:28:12 -0800, Jeff Liebermann <jeffl@cruzio.com>
    wrote:

    On Mon, 13 Jan 2025 08:57:18 -0500, zen cycle
    <funkmasterxx@hotmail.com> wrote:

    On 1/13/2025 2:01 AM, Jeff Liebermann wrote:
    On Mon, 13 Jan 2025 13:20:04 +0700, John B. <slocombjb@gmail.com>
    wrote:

    But long before the days of transistors and "circuit boards" my folks >>>>> had an electric blanket... used it all the rest of their lives. Is the >>>>> modern one better?

    Not really better, but certainly safer. Electric heaters are all 100% >>>> efficient. Every watt you shove into the heating wires is turned into >>>> heat. There's no way to improve on 100% efficiency.

    Nothing is 100% efficient

    Well, the heater wires might radiate some EM radiation, which then is >>absorbed by nearby objects and is converted to IR which adds to the
    heat produced by the heater wires. There might be some radiation at
    other frequencies (RFI, EMI, microwaves, visible light, UV, etc) but
    most of the radiation is ends up somewhere in the IR bands.

    Incidentally, I'm usually amused at the advertising for electric
    heaters all claiming that their more "efficient" than the competition.
    Of course, no numbers and certainly no calculations are ever provided.

    They could always do a study where they ask people which heater made
    them feel warmer.

    I've seen that in advertisements and ad agency generated "scientific"
    studies. Usually, the winner is the electric heater with a parabolic
    reflector to concentrate the heat in a small area. The dish is not a
    true parabola. If it were, such a heater could probably burn a hole
    in whatever it's pointed at. The local Costco store usually has a
    working model on display when winter weather arrives. I would walk
    down the isle and instantly feel the heat. <https://www.homedepot.com/b/Heating-Venting-Cooling-Heaters-Space-Heaters/Dish/N-5yc1vZc4lhZ1z0wblh>







    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to All on Thu Jan 16 21:25:38 2025
    On Wed, 15 Jan 2025 15:03:44 -0000 (UTC), Ted Heise <theise@panix.com>
    wrote:

    On Mon, 13 Jan 2025 21:13:57 -0800,
    Jeff Liebermann <jeffl@cruzio.com> wrote:
    On Mon, 13 Jan 2025 12:58:35 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:
    On 1/13/2025 12:06 PM, Wolfgang Strobl wrote:
    You, as a person, don't need or consume heat. Somewhat
    simplified, you need a certain range of temperatures. In the
    ideal case, you don't need any additional energy, because
    your body already produces heat. A little bit of isolation,
    perhaps provided by that very blanket, might be sufficient.

    And, I suppose, we could do away with all home heating, and
    just wear very heavy clothing all winter. But I don't know of
    anyone doing that.

    I'm currently doing something similar to save firewood, which
    costs about $500/cord delivered.

    Oh my gosh, that's outrageous! A buddy and I cut firewood for a
    living in Flagstaff one summer. We charged $30 for a cord of pine
    (cut, split, delivered and stacked), Oak was $40.

    When I first moved to Santa Cruz in about 1974, dry oak and madrone
    firewood was about $50 delivered. Of course, the vendors always
    provided less than one cord and not so dry (green) wood which required
    an extra year to dry out. Nobody wanted pine, which burns fast and
    hot. There are also idiots that burn wet pine which generates lots of creosote. However, the real problem is that the energy generated
    (BTU/cord) for pine (16 million BTU/cord) is much lower than oak (29
    million BTU/cord. I burned some pine long ago, and never repeated
    that mistake.

    <https://modernsurvivalblog.com/alternative-energy/best-wood-for-heating/> "...a cord of Pine will only provide 60% of the heat as Oak"

    My wood pile 2024-2025:
    <https://photos.app.goo.gl/hRQZHVcHtykygudz9>

    My wood piles over the years: <https://www.learnbydestroying.com/jeffl/pics/home/>

    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Roger Merriman@21:1/5 to Frank Krygowski on Fri Jan 17 15:13:41 2025
    Frank Krygowski <frkrygow@sbcglobal.net> wrote:
    On 1/16/2025 4:21 AM, Catrike Ryder wrote:
    On Wed, 15 Jan 2025 21:05:22 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    IOW, we are not allowed to discuss reliability of bike parts?

    1) Usage by a half dozen or so of Roger's friends doesn't even begin
    to define reliability.
    Gosh, does that mean we should look at actual large population _data_????

    Why, that can't be right! It goes against half the arguments you've ever
    made here! :-)


    Considering the increase in complexity and that electronic components can
    and do fail it adds another potential failure point though I believe some
    of the newer SRAM rear mech designs and maybe Shimano? Are more tolerant of rock strikes, which some MTBers can be rather prone to ie it’s likely to
    die by impact than electronic components failing.

    Roger Merriman

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jeff Liebermann@21:1/5 to frkrygow@sbcglobal.net on Fri Jan 17 10:29:17 2025
    On Fri, 17 Jan 2025 12:06:29 -0500, Frank Krygowski
    <frkrygow@sbcglobal.net> wrote:

    On 1/17/2025 12:33 AM, Jeff Liebermann wrote:
    On Mon, 13 Jan 2025 14:33:13 -0500, Catrike Ryder
    <Soloman@old.bikers.org> wrote:

    On Mon, 13 Jan 2025 11:28:12 -0800, Jeff Liebermann <jeffl@cruzio.com>
    wrote:

    On Mon, 13 Jan 2025 08:57:18 -0500, zen cycle
    <funkmasterxx@hotmail.com> wrote:

    On 1/13/2025 2:01 AM, Jeff Liebermann wrote:
    On Mon, 13 Jan 2025 13:20:04 +0700, John B. <slocombjb@gmail.com>
    wrote:

    But long before the days of transistors and "circuit boards" my folks >>>>>>> had an electric blanket... used it all the rest of their lives. Is the >>>>>>> modern one better?

    Not really better, but certainly safer. Electric heaters are all 100% >>>>>> efficient. Every watt you shove into the heating wires is turned into >>>>>> heat. There's no way to improve on 100% efficiency.

    Nothing is 100% efficient

    Well, the heater wires might radiate some EM radiation, which then is
    absorbed by nearby objects and is converted to IR which adds to the
    heat produced by the heater wires. There might be some radiation at
    other frequencies (RFI, EMI, microwaves, visible light, UV, etc) but
    most of the radiation is ends up somewhere in the IR bands.

    Incidentally, I'm usually amused at the advertising for electric
    heaters all claiming that their more "efficient" than the competition. >>>> Of course, no numbers and certainly no calculations are ever provided.

    They could always do a study where they ask people which heater made
    them feel warmer.

    I've seen that in advertisements and ad agency generated "scientific"
    studies. Usually, the winner is the electric heater with a parabolic
    reflector to concentrate the heat in a small area. The dish is not a
    true parabola. If it were, such a heater could probably burn a hole
    in whatever it's pointed at.

    A mathematical quibble: You may be thinking of an ellipsoidal reflector.
    An ellipsoid would take the radiation emanating from its focal point and >concentrate it at its other focal point. And at other locations, there >wouldn't be much concentration anyway.

    A parabolic (or parabaloid) reflector would take the radiation emanating
    from its focal point and send it out in parallel lines. Again, no great >concentration.

    See >https://sciencetech-inc.com/web/content/?model=product.template.website_pdf&id=1297&filename_field=name&field=doc&filename=doc.pdf

    Nope. I used the term "parabolic dish" to describe a directional
    reflector which recognizes. I don't know exactly the shape of the
    reflector in the heater, but I do know that with such small f/D (focal
    length to diameter ratio) and very wide heating element, the reflector
    could be almost any shape and work much in the same way.

    The design of such a directional bathroom heater has some goals:
    1. Spreading the IR radiation over an area that approximates the
    width of a human.
    2. NOT concentrating the IR radiation to produce hot spots that might
    injure the user or start a fire.
    3. Convince UL that it's safe and won't turn into a death ray.

    Argh... I need to go somewhere. I'll fill in the tech stuff when I
    return.


    --
    Jeff Liebermann jeffl@cruzio.com
    PO Box 272 http://www.LearnByDestroying.com
    Ben Lomond CA 95005-0272
    Skype: JeffLiebermann AE6KS 831-336-2558

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Zen Cycle@21:1/5 to Radey Shouman on Fri Jan 17 15:58:54 2025
    On 1/16/2025 5:56 PM, Radey Shouman wrote:
    Zen Cycle <funkmaster@hotmail.com> writes:

    On 1/16/2025 12:18 PM, Frank Krygowski wrote:
    On 1/16/2025 7:14 AM, zen cycle wrote:
    On 1/15/2025 6:39 PM, Frank Krygowski wrote:

    You should look at the energy used for the controls and think
    about what becomes of it. Do that in microcopic detail.

    If you think it all ends up as heat (IR spectrum) you have a gross
    misunderstanding of electronics. First off, the indicators
    dissipate energy in the visible light spectrum (this is why LEDs
    are more efficient lighting than incandescent, very little energy
    is used in the IR spectrum).
    OK, a thought experiment: Take an adiabatic container - that is, a
    _perfectly_ insulated box (a physical impossibility, but useful for
    our analysis). Let the box contain whatever you like - just air,
    some solid objects, whatever.
    Cut an LED sized hole in it and insert an LED of your choice so it
    shines into the box. Turn on the LED.
    What happens to the light entering the box? Obviously, you don't end
    up with a box full of light, so it isn't stored; it somehow goes
    away.

    lol...no, it doesn't 'go away'. For all intents and purposes it
    suffers the safe fate as Schrödinger's cat.

    And what happens to the temperature inside the box, and why?

    It will rise somewhat due to the residual IR energy from the
    system. Visible spectrum LEDs do emit some IR, just not nearly as much
    as incandescents (given the same lux)

    Answer: The temperature of whatever's inside the box will rise. The
    energy put into _all_ spectra by the LED, including the visible
    light spectrum, ultimately converts to heat.

    So _all_ forms of energy convert to heat? You should write a paper on that. >>

    Secondly, think about your premise that it all turns into heat -
    this means no energy is available to do any other form of work.
    In physics or mechanical engineering, work is defined as force
    moving through a distance, or torque moving though an angle of
    rotation. Valid units of measurement are the same as the units for
    energy: foot*pounds, Newton*meters or Joules, etc. all of which
    (interestingly) can be converted to BTUs, which are normally units
    measuring heat.
    And in general, you're right, energy converted to heat is not
    normally available to do work.

    I used the term 'work' more generically, in this case running a
    program, setting bits in memory, etc. Yes, residual heat from the
    process, but energy is used to perform whatever task, Heat is the
    result of losses in the system (thermal junctions from die bonds, for
    example)

    I don't believe that is correct. All the energy used to run a program
    does eventually get rejected as heat. I suppose either 1 or 0 bits must
    have a slightly higher potential energy, but the net number of each is
    not likely to change much, and the energy difference must be small.

    Hmmm, so you're suggesting that if we consider computers to be heaters
    that also perform a computing function, that using it as a heater
    ostensibly gets us computing functions for free as long as we're
    expecting the heating function to be primary?

    Sounds suspiciously perpetual-motion-esque.


    There is a field of study called "thermodynamic computing", which
    studies the minimum energy that must be dissipated for various logic operations. Turns out that reversible computations, ie those that do
    not irretrievably lose information, can be more thermodynamically
    efficient than those that do lose information. Eventually this sort of thinking is hoped to enable more economical computation.

    RAM vs NVRAM? I think that's been done before.


    --
    Add xx to reply

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Radey Shouman@21:1/5 to Zen Cycle on Fri Jan 17 17:13:01 2025
    Zen Cycle <funkmaster@hotmail.com> writes:

    On 1/16/2025 5:56 PM, Radey Shouman wrote:
    Zen Cycle <funkmaster@hotmail.com> writes:

    On 1/16/2025 12:18 PM, Frank Krygowski wrote:
    On 1/16/2025 7:14 AM, zen cycle wrote:
    On 1/15/2025 6:39 PM, Frank Krygowski wrote:

    You should look at the energy used for the controls and think
    about what becomes of it. Do that in microcopic detail.

    If you think it all ends up as heat (IR spectrum) you have a gross
    misunderstanding of electronics. First off, the indicators
    dissipate energy in the visible light spectrum (this is why LEDs
    are more efficient lighting than incandescent, very little energy
    is used in the IR spectrum).
    OK, a thought experiment: Take an adiabatic container - that is, a
    _perfectly_ insulated box (a physical impossibility, but useful for
    our analysis). Let the box contain whatever you like - just air,
    some solid objects, whatever.
    Cut an LED sized hole in it and insert an LED of your choice so it
    shines into the box. Turn on the LED.
    What happens to the light entering the box? Obviously, you don't end
    up with a box full of light, so it isn't stored; it somehow goes
    away.

    lol...no, it doesn't 'go away'. For all intents and purposes it
    suffers the safe fate as Schrödinger's cat.

    And what happens to the temperature inside the box, and why?

    It will rise somewhat due to the residual IR energy from the
    system. Visible spectrum LEDs do emit some IR, just not nearly as much
    as incandescents (given the same lux)

    Answer: The temperature of whatever's inside the box will rise. The
    energy put into _all_ spectra by the LED, including the visible
    light spectrum, ultimately converts to heat.

    So _all_ forms of energy convert to heat? You should write a paper on that. >>>

    Secondly, think about your premise that it all turns into heat -
    this means no energy is available to do any other form of work.
    In physics or mechanical engineering, work is defined as force
    moving through a distance, or torque moving though an angle of
    rotation. Valid units of measurement are the same as the units for
    energy: foot*pounds, Newton*meters or Joules, etc. all of which
    (interestingly) can be converted to BTUs, which are normally units
    measuring heat.
    And in general, you're right, energy converted to heat is not
    normally available to do work.

    I used the term 'work' more generically, in this case running a
    program, setting bits in memory, etc. Yes, residual heat from the
    process, but energy is used to perform whatever task, Heat is the
    result of losses in the system (thermal junctions from die bonds, for
    example)
    I don't believe that is correct. All the energy used to run a
    program
    does eventually get rejected as heat. I suppose either 1 or 0 bits must
    have a slightly higher potential energy, but the net number of each is
    not likely to change much, and the energy difference must be small.

    Hmmm, so you're suggesting that if we consider computers to be heaters
    that also perform a computing function, that using it as a heater
    ostensibly gets us computing functions for free as long as we're
    expecting the heating function to be primary?

    Sounds suspiciously perpetual-motion-esque.

    There's nothing of perpetual motion about it. Essentially all of the
    power you put into a computer is dissipated as heat. Even mechanical
    work, like driving a computer fan, fairly quickly results in heating the
    room. A small amount of energy leaves the computer as EM radiation,
    visible if you include the monitor, and intentional and unintentional RF transmission and that's about it.

    If you don't believe this, tell me what electrical phenomenon consumes
    energy without producing either radiation, mechanical work, or heat.
    All of the energy used for useful computing is lost by passing currents
    through components with a voltage drop -- resistors, diodes, and
    transistors. All dissipate the energy used as heat.

    There is a field of study called "thermodynamic computing", which
    studies the minimum energy that must be dissipated for various logic
    operations. Turns out that reversible computations, ie those that do
    not irretrievably lose information, can be more thermodynamically
    efficient than those that do lose information. Eventually this sort of
    thinking is hoped to enable more economical computation.

    RAM vs NVRAM? I think that's been done before.

    I'm afraid you misunderstand, but I don't want to get into a rathole
    that will amuse almost none of our readers.

    --

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Zen Cycle on Fri Jan 17 16:14:30 2025
    On 1/17/2025 2:58 PM, Zen Cycle wrote:
    On 1/16/2025 5:56 PM, Radey Shouman wrote:
    Zen Cycle <funkmaster@hotmail.com> writes:

    On 1/16/2025 12:18 PM, Frank Krygowski wrote:
    On 1/16/2025 7:14 AM, zen cycle wrote:
    On 1/15/2025 6:39 PM, Frank Krygowski wrote:

    You should look at the energy used for the controls
    and think
    about what becomes of it. Do that in microcopic detail.

    If you think it all ends up as heat (IR spectrum) you
    have a gross
    misunderstanding of electronics. First off, the indicators
    dissipate energy in the visible light spectrum (this is
    why LEDs
    are more efficient lighting than incandescent, very
    little energy
    is used in the IR spectrum).
    OK, a thought experiment: Take an adiabatic container -
    that is, a
    _perfectly_ insulated box (a physical impossibility, but
    useful for
    our analysis). Let the box contain whatever you like -
    just air,
    some solid objects, whatever.
    Cut an LED sized hole in it and insert an LED of your
    choice so it
    shines into the box. Turn on the LED.
    What happens to the light entering the box? Obviously,
    you don't end
    up with a box full of light, so it isn't stored; it
    somehow goes
    away.

    lol...no, it doesn't 'go away'. For all intents and
    purposes it
    suffers the safe fate as Schrödinger's cat.

    And what happens to the temperature inside the box, and
    why?

    It will rise somewhat due to the residual IR energy from the
    system. Visible spectrum LEDs do emit some IR, just not
    nearly as much
    as incandescents (given the same lux)

    Answer: The temperature of whatever's inside the box
    will rise. The
    energy put into _all_ spectra by the LED, including the
    visible
    light spectrum, ultimately converts to heat.

    So _all_ forms of energy convert to heat? You should
    write a paper on that.


    Secondly, think about your premise that it all turns
    into heat -
    this means no energy is available to do any other form
    of work.
    In physics or mechanical engineering, work is defined as
    force
    moving through a distance, or torque moving though an
    angle of
    rotation. Valid units of measurement are the same as the
    units for
    energy: foot*pounds, Newton*meters or Joules, etc. all
    of which
    (interestingly) can be converted to BTUs, which are
    normally units
    measuring heat.
    And in general, you're right, energy converted to heat
    is not
    normally available to do work.

    I used the term 'work' more generically, in this case
    running a
    program, setting bits in memory, etc. Yes, residual heat
    from the
    process, but energy is used to perform whatever task,
    Heat is the
    result of losses in the system (thermal junctions from
    die bonds, for
    example)

    I don't believe that is correct.  All the energy used to
    run a program
    does eventually get rejected as heat.  I suppose either 1
    or 0 bits must
    have a slightly higher potential energy, but the net
    number of each is
    not likely to change much, and the energy difference must
    be small.

    Hmmm, so you're suggesting that if we consider computers to
    be heaters that also perform a computing function, that
    using it as a heater ostensibly gets us computing functions
    for free as long as we're expecting the heating function to
    be primary?

    Sounds suspiciously perpetual-motion-esque.


    There is a field of study called "thermodynamic
    computing", which
    studies the minimum energy that must be dissipated for
    various logic
    operations.  Turns out that reversible computations, ie
    those that do
    not irretrievably lose information, can be more
    thermodynamically
    efficient than those that do lose information.  Eventually
    this sort of
    thinking is hoped to enable more economical computation.

    RAM vs NVRAM? I think that's been done before.




    Nice. But all those arguments on 'intent' are not about
    actual physics

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Radey Shouman on Fri Jan 17 16:35:35 2025
    On 1/17/2025 4:13 PM, Radey Shouman wrote:
    Zen Cycle <funkmaster@hotmail.com> writes:

    On 1/16/2025 5:56 PM, Radey Shouman wrote:
    Zen Cycle <funkmaster@hotmail.com> writes:

    On 1/16/2025 12:18 PM, Frank Krygowski wrote:
    On 1/16/2025 7:14 AM, zen cycle wrote:
    On 1/15/2025 6:39 PM, Frank Krygowski wrote:

    You should look at the energy used for the controls and think
    about what becomes of it. Do that in microcopic detail.

    If you think it all ends up as heat (IR spectrum) you have a gross >>>>>> misunderstanding of electronics. First off, the indicators
    dissipate energy in the visible light spectrum (this is why LEDs
    are more efficient lighting than incandescent, very little energy
    is used in the IR spectrum).
    OK, a thought experiment: Take an adiabatic container - that is, a
    _perfectly_ insulated box (a physical impossibility, but useful for
    our analysis). Let the box contain whatever you like - just air,
    some solid objects, whatever.
    Cut an LED sized hole in it and insert an LED of your choice so it
    shines into the box. Turn on the LED.
    What happens to the light entering the box? Obviously, you don't end >>>>> up with a box full of light, so it isn't stored; it somehow goes
    away.

    lol...no, it doesn't 'go away'. For all intents and purposes it
    suffers the safe fate as Schrödinger's cat.

    And what happens to the temperature inside the box, and why?

    It will rise somewhat due to the residual IR energy from the
    system. Visible spectrum LEDs do emit some IR, just not nearly as much >>>> as incandescents (given the same lux)

    Answer: The temperature of whatever's inside the box will rise. The
    energy put into _all_ spectra by the LED, including the visible
    light spectrum, ultimately converts to heat.

    So _all_ forms of energy convert to heat? You should write a paper on that.


    Secondly, think about your premise that it all turns into heat -
    this means no energy is available to do any other form of work.
    In physics or mechanical engineering, work is defined as force
    moving through a distance, or torque moving though an angle of
    rotation. Valid units of measurement are the same as the units for
    energy: foot*pounds, Newton*meters or Joules, etc. all of which
    (interestingly) can be converted to BTUs, which are normally units
    measuring heat.
    And in general, you're right, energy converted to heat is not
    normally available to do work.

    I used the term 'work' more generically, in this case running a
    program, setting bits in memory, etc. Yes, residual heat from the
    process, but energy is used to perform whatever task, Heat is the
    result of losses in the system (thermal junctions from die bonds, for
    example)
    I don't believe that is correct. All the energy used to run a
    program
    does eventually get rejected as heat. I suppose either 1 or 0 bits must >>> have a slightly higher potential energy, but the net number of each is
    not likely to change much, and the energy difference must be small.

    Hmmm, so you're suggesting that if we consider computers to be heaters
    that also perform a computing function, that using it as a heater
    ostensibly gets us computing functions for free as long as we're
    expecting the heating function to be primary?

    Sounds suspiciously perpetual-motion-esque.

    There's nothing of perpetual motion about it. Essentially all of the
    power you put into a computer is dissipated as heat. Even mechanical
    work, like driving a computer fan, fairly quickly results in heating the room. A small amount of energy leaves the computer as EM radiation,
    visible if you include the monitor, and intentional and unintentional RF transmission and that's about it.

    If you don't believe this, tell me what electrical phenomenon consumes
    energy without producing either radiation, mechanical work, or heat.
    All of the energy used for useful computing is lost by passing currents through components with a voltage drop -- resistors, diodes, and
    transistors. All dissipate the energy used as heat.

    There is a field of study called "thermodynamic computing", which
    studies the minimum energy that must be dissipated for various logic
    operations. Turns out that reversible computations, ie those that do
    not irretrievably lose information, can be more thermodynamically
    efficient than those that do lose information. Eventually this sort of
    thinking is hoped to enable more economical computation.

    RAM vs NVRAM? I think that's been done before.

    I'm afraid you misunderstand, but I don't want to get into a rathole
    that will amuse almost none of our readers.


    Theoretical physicists discuss information itself as if it
    were a real thing. They lose me along the way.

    https://arxiv.org/pdf/1507.03131

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Wolfgang Strobl@21:1/5 to All on Sun Jan 19 11:37:58 2025
    Am Wed, 15 Jan 2025 15:42:39 -0500 schrieb Zen Cycle
    <funkmaster@hotmail.com>:

    On 1/15/2025 1:42 PM, Frank Krygowski wrote:
    On 1/15/2025 1:28 PM, Zen Cycle wrote:
    On 1/15/2025 1:16 PM, Frank Krygowski wrote:
    On 1/15/2025 1:05 PM, Zen Cycle wrote:
    On 1/13/2025 11:03 AM, Frank Krygowski wrote:

    It's certainly true that 100% of the electricity consumed by an
    electric blanket becomes heat.

    No, that isn't true either.

    Please explain. What electrical energy goes elsewhere?

    A very small amount of power is used for the indicator lighting and
    electronic controls.


    I write "either" because even _if_ it were true that electric
    heaters are 100% efficient (which isn't true), saying 100% of the
    electricity consumed by the device become heat is very different
    than saying it's 100% efficient.

    What's your definition of "efficiency?" As I said earlier, I think a
    common one used for engineering matters is Desired Output divided by
    Required Input, or something similar.

    Do you have a different one?

    Nope, it's the notion that every watt of power directly goes into
    heating the targeted space that I'm stuck on.

    You're moving goalposts. You objected to my statement "It's certainly
    true that 100% of the electricity consumed by an electric blanket
    becomes heat."

    Nope, I addressed that point very specifically with "A very small amount
    of power is used for the indicator lighting and electronic controls."

    What you misinterpreted as 'moving the goalposts' was me taking issue
    with Jeffs assertion that "electric heaters are all 100% efficient".

    Note that "electric heaters are all 100% efficient" ? "It's certainly
    true that 100% of the electricity consumed by an electric blanket
    becomes heat."

    The efficiency of the heater is determined by the energy that is used >specifically for generating heat. By that premise, it's logically
    possible that that the heating element in a heating appliance may be
    near 100%, but that some energy will be used for the control portion of
    the system.


    Provided that the “control part” of the heater is not located in a
    remote data center, this is completely irrelevant. If the control system
    is part of the heater, the electrical energy used by the control is also converted into heat.


    If Jeff had written "Electric heaters are all _nearly_ 100% efficient" I >wouldn't have commented.

    You are barking under the wrong tree.

    Of course an electric blanket isn't 100% efficient. But this doesn't
    have anything to do with how electricity is converted to heat.

    The relevant question is how much of the electrical energy supplied to
    an isolated system serves the intended purpose, and where we draw the
    line. If an electric blanket is intended to warm a person, but eighty
    percent of the heat only slightly warms the cold air above the blanket
    or warms the walls through radiation, then the efficiency of the
    electric blanket is only twenty percent. This does not change the fact
    that the loss-free conversion of electricity into heat through an
    resistor is just that: loss-free.

    Arguing that some heat might be stored inside the insulation of the
    cable from the wall socket to the blanket for a few minutes is just nit-picking. There is no point to insert various "almost", "nearly",
    "in the limit" into valid generalizations.

    I don't doubt that "electric heaters are all 100% efficient" is
    misleading. But not because of those tiny losses you mention.

    There is a popular misconception that this physical triviality can be
    applied to the entire system, from the original energy source to the
    user of the heat generated.

    No energy source delivers even close to 100%. Even a solar panel has to
    be produced and replaced every twenty or thirty years. There are energy requirements for production, maintenance and recycling of power lines. Renewables are much better than the alternatives, but still nowhere near
    100%.


    --
    Thank you for observing all safety precautions

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From zen cycle@21:1/5 to Radey Shouman on Sun Jan 19 08:29:31 2025
    On 1/17/2025 5:13 PM, Radey Shouman wrote:
    Zen Cycle <funkmaster@hotmail.com> writes:

    On 1/16/2025 5:56 PM, Radey Shouman wrote:
    Zen Cycle <funkmaster@hotmail.com> writes:

    On 1/16/2025 12:18 PM, Frank Krygowski wrote:
    On 1/16/2025 7:14 AM, zen cycle wrote:
    On 1/15/2025 6:39 PM, Frank Krygowski wrote:

    You should look at the energy used for the controls and think
    about what becomes of it. Do that in microcopic detail.

    If you think it all ends up as heat (IR spectrum) you have a gross >>>>>> misunderstanding of electronics. First off, the indicators
    dissipate energy in the visible light spectrum (this is why LEDs
    are more efficient lighting than incandescent, very little energy
    is used in the IR spectrum).
    OK, a thought experiment: Take an adiabatic container - that is, a
    _perfectly_ insulated box (a physical impossibility, but useful for
    our analysis). Let the box contain whatever you like - just air,
    some solid objects, whatever.
    Cut an LED sized hole in it and insert an LED of your choice so it
    shines into the box. Turn on the LED.
    What happens to the light entering the box? Obviously, you don't end >>>>> up with a box full of light, so it isn't stored; it somehow goes
    away.

    lol...no, it doesn't 'go away'. For all intents and purposes it
    suffers the safe fate as Schrödinger's cat.

    And what happens to the temperature inside the box, and why?

    It will rise somewhat due to the residual IR energy from the
    system. Visible spectrum LEDs do emit some IR, just not nearly as much >>>> as incandescents (given the same lux)

    Answer: The temperature of whatever's inside the box will rise. The
    energy put into _all_ spectra by the LED, including the visible
    light spectrum, ultimately converts to heat.

    So _all_ forms of energy convert to heat? You should write a paper on that.


    Secondly, think about your premise that it all turns into heat -
    this means no energy is available to do any other form of work.
    In physics or mechanical engineering, work is defined as force
    moving through a distance, or torque moving though an angle of
    rotation. Valid units of measurement are the same as the units for
    energy: foot*pounds, Newton*meters or Joules, etc. all of which
    (interestingly) can be converted to BTUs, which are normally units
    measuring heat.
    And in general, you're right, energy converted to heat is not
    normally available to do work.

    I used the term 'work' more generically, in this case running a
    program, setting bits in memory, etc. Yes, residual heat from the
    process, but energy is used to perform whatever task, Heat is the
    result of losses in the system (thermal junctions from die bonds, for
    example)
    I don't believe that is correct. All the energy used to run a
    program
    does eventually get rejected as heat. I suppose either 1 or 0 bits must >>> have a slightly higher potential energy, but the net number of each is
    not likely to change much, and the energy difference must be small.

    Hmmm, so you're suggesting that if we consider computers to be heaters
    that also perform a computing function, that using it as a heater
    ostensibly gets us computing functions for free as long as we're
    expecting the heating function to be primary?

    Sounds suspiciously perpetual-motion-esque.

    There's nothing of perpetual motion about it. Essentially all of the
    power you put into a computer is dissipated as heat.

    "essentially".....thank you.

    'essentially', 'virtually', 'practically', Feel free to you any limiting adjverb you like, it's not 100%, this is been my point all along.


    Even mechanical
    work, like driving a computer fan, fairly quickly results in heating the room. A small amount of energy leaves the computer as EM radiation,
    visible if you include the monitor, and intentional and unintentional RF transmission and that's about it.

    If you don't believe this, tell me what electrical phenomenon consumes
    energy without producing either radiation, mechanical work, or heat.
    All of the energy used for useful computing is lost by passing currents through components with a voltage drop -- resistors, diodes, and
    transistors. All dissipate the energy used as heat.

    There is a field of study called "thermodynamic computing", which
    studies the minimum energy that must be dissipated for various logic
    operations. Turns out that reversible computations, ie those that do
    not irretrievably lose information, can be more thermodynamically
    efficient than those that do lose information. Eventually this sort of
    thinking is hoped to enable more economical computation.

    RAM vs NVRAM? I think that's been done before.

    I'm afraid you misunderstand, but I don't want to get into a rathole
    that will amuse almost none of our readers.


    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From AMuzi@21:1/5 to Wolfgang Strobl on Sun Jan 19 09:05:32 2025
    On 1/19/2025 4:37 AM, Wolfgang Strobl wrote:
    Am Wed, 15 Jan 2025 15:42:39 -0500 schrieb Zen Cycle <funkmaster@hotmail.com>:

    On 1/15/2025 1:42 PM, Frank Krygowski wrote:
    On 1/15/2025 1:28 PM, Zen Cycle wrote:
    On 1/15/2025 1:16 PM, Frank Krygowski wrote:
    On 1/15/2025 1:05 PM, Zen Cycle wrote:
    On 1/13/2025 11:03 AM, Frank Krygowski wrote:

    It's certainly true that 100% of the electricity consumed by an
    electric blanket becomes heat.

    No, that isn't true either.

    Please explain. What electrical energy goes elsewhere?

    A very small amount of power is used for the indicator lighting and
    electronic controls.


    I write "either" because even _if_ it were true that electric
    heaters are 100% efficient (which isn't true), saying 100% of the
    electricity consumed by the device become heat is very different
    than saying it's 100% efficient.

    What's your definition of "efficiency?" As I said earlier, I think a >>>>> common one used for engineering matters is Desired Output divided by >>>>> Required Input, or something similar.

    Do you have a different one?

    Nope, it's the notion that every watt of power directly goes into
    heating the targeted space that I'm stuck on.

    You're moving goalposts. You objected to my statement "It's certainly
    true that 100% of the electricity consumed by an electric blanket
    becomes heat."

    Nope, I addressed that point very specifically with "A very small amount
    of power is used for the indicator lighting and electronic controls."

    What you misinterpreted as 'moving the goalposts' was me taking issue
    with Jeffs assertion that "electric heaters are all 100% efficient".

    Note that "electric heaters are all 100% efficient" ? "It's certainly
    true that 100% of the electricity consumed by an electric blanket
    becomes heat."

    The efficiency of the heater is determined by the energy that is used
    specifically for generating heat. By that premise, it's logically
    possible that that the heating element in a heating appliance may be
    near 100%, but that some energy will be used for the control portion of
    the system.


    Provided that the “control part” of the heater is not located in a
    remote data center, this is completely irrelevant. If the control system
    is part of the heater, the electrical energy used by the control is also converted into heat.


    If Jeff had written "Electric heaters are all _nearly_ 100% efficient" I
    wouldn't have commented.

    You are barking under the wrong tree.

    Of course an electric blanket isn't 100% efficient. But this doesn't
    have anything to do with how electricity is converted to heat.

    The relevant question is how much of the electrical energy supplied to
    an isolated system serves the intended purpose, and where we draw the
    line. If an electric blanket is intended to warm a person, but eighty
    percent of the heat only slightly warms the cold air above the blanket
    or warms the walls through radiation, then the efficiency of the
    electric blanket is only twenty percent. This does not change the fact
    that the loss-free conversion of electricity into heat through an
    resistor is just that: loss-free.

    Arguing that some heat might be stored inside the insulation of the
    cable from the wall socket to the blanket for a few minutes is just nit-picking. There is no point to insert various "almost", "nearly",
    "in the limit" into valid generalizations.

    I don't doubt that "electric heaters are all 100% efficient" is
    misleading. But not because of those tiny losses you mention.

    There is a popular misconception that this physical triviality can be
    applied to the entire system, from the original energy source to the
    user of the heat generated.

    No energy source delivers even close to 100%. Even a solar panel has to
    be produced and replaced every twenty or thirty years. There are energy requirements for production, maintenance and recycling of power lines. Renewables are much better than the alternatives, but still nowhere near 100%.



    +1
    The gap is between physics and epistemology of the specific
    example.

    --
    Andrew Muzi
    am@yellowjersey.org
    Open every day since 1 April, 1971

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)