My PC takes ages (up to 10 minutes) to boot from "cold" (measuring until
all the background system tray applets have started and disk-thrashing
ceases, so most days I tend to put it to sleep rather than shutting it down.

I've noticed recently that it often (but not always) fails to start up
from sleep mode if a USB hard disk is connected as I press the power
soft-start switch. The hard disk light comes on continuously, as if the
memory contents are being restored from the HDD copy that sleep-mode
uses, but the graphics card does not turn on (no logon screen or
anything else displayed) and after a few seconds the power light reverts
to its orange "asleep" state. It's as if it wakes up momentarily and
then almost immediately goes back to sleep.

Is this likely to be a Windows issue or a BIOS/UEFI issue?

I've not yet established whether the problem is confined to one specific external HDD or whether other disks cause the same symptoms.

It's not a problem: I just need to remember not to plug the backup drive
in until I see a logo screen to prove that the PC is awake.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On Mon, 5/12/2025 6:14 PM, NY wrote:

My PC takes ages (up to 10 minutes) to boot from "cold" (measuring until all the background system tray applets have started and disk-thrashing ceases, so most days I tend to put it to sleep rather than shutting it down.

I've noticed recently that it often (but not always) fails to start up from sleep mode if a USB hard disk is connected as I press the power soft-start switch. The hard disk light comes on continuously, as if the memory contents are being restored from

the HDD copy that sleep-mode uses, but the graphics card does not turn on (no logon screen or anything else displayed) and after a few seconds the power light reverts to its orange "asleep" state. It's as if it wakes up momentarily and then almost
immediately goes back to sleep.

Is this likely to be a Windows issue or a BIOS/UEFI issue?

I've not yet established whether the problem is confined to one specific external HDD or whether other disks cause the same symptoms.

It's not a problem: I just need to remember not to plug the backup drive in until I see a logo screen to prove that the PC is awake.

Check the power supply label, the one with the power ratings.

The USB HDD is drawing from the +5VSB supply.

That same supply is used for supervisor operation, and
the +5VSB must be operational, for the soft power button
on the front of the PC to work.

All it would take, is plugging in an Apple Device that wants
2 amps of current, to affect the ability of the PC to come
back to life. Some of the various Apples can draw +5VSB @ 2A
while charging.

The PC design wasn't always this way. At one time, you could set
some of the ports to run off +5VSB and some to run off the
main +5V (closer to 20 amp capability). This allowed the user
to steer the heavier loads to a better choice of powering.
But the PC makers could not be happy with a scheme that
worked, so we have the potential for various power-starved
scenarios.

On an older ATX PSU, with an inefficient +5VSB, while the PC
sleeps, the +5VSB supply gets hot. When you start the PC, there
is a "puff of hot air" from the PSU vent, and this is the
heat generated while it was asleep and fending off a too-large load.
I have an old machine that does this. Some of the +5VSB were linear
regulators, which are extremely inefficient. Most of them today
are SMPS for the "main supply" and SMPS for the +5VSB supervisor supply.
It makes less waste heat and lowers the power bill per year.

You'll have to estimate the loading, to see if any of this makes sense.

Paul

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

NY <me@privacy.net> wrote:

My PC takes ages (up to 10 minutes) to boot from "cold" (measuring until
all the background system tray applets have started and disk-thrashing ceases, so most days I tend to put it to sleep rather than shutting it down.

I've noticed recently that it often (but not always) fails to start up
from sleep mode if a USB hard disk is connected as I press the power soft-start switch. The hard disk light comes on continuously, as if the memory contents are being restored from the HDD copy that sleep-mode
uses, but the graphics card does not turn on (no logon screen or
anything else displayed) and after a few seconds the power light reverts
to its orange "asleep" state. It's as if it wakes up momentarily and
then almost immediately goes back to sleep.

Are you sure it's sleeping? What you describe is hibernation.

When sleeping, you don't have to use "the power soft-start switch",
but can just wake up the computer with a keyboard press or 'mouse'
movement.

Also when sleeping, there is no situation of "memory contents are
being restored from the HDD copy that sleep-mode uses", because memory
contents are preserved in RAM when sleeping. (Yes, there is some mode
when memory contents can also be copied to HDD, just in case the system
loses power when sleeping. I forgot what that mode is called and how
it's set. 'Hybrid Sleep'? (If so, I don't have Hybrid Sleep, because I
have a laptop, which doesn't need that mode.))

If the system is really sleeping (a normal sleep), it might be worth
your while to set/tick the 'Allow the computer to turn off this device
to save power' setting on the 'Power Management' tab of the (correct)
'USB Mass Strage Device' in the 'Universal Seerial Bus Controllers' tree
of Device Manager. Maybe with that setting, Windows might not touch the
USB HDD till the rest of the system has been woken up (sufficiently).

Is this likely to be a Windows issue or a BIOS/UEFI issue?

AFAIK, when sleeping, the BIOS/UEFI is not involved, because Windows
is still in charge [1]. But Paul would be better informed in this area.

I've not yet established whether the problem is confined to one specific external HDD or whether other disks cause the same symptoms.

In any case, always use the *same* USB port, because the 'Allow the
computer to turn off...' setting is per-port.

It's not a problem: I just need to remember not to plug the backup drive
in until I see a logo screen to prove that the PC is awake.

[1] If Windows isn't in control, it couldn't be configured (for example
with powercfg) to set which devices can and can not wake up the system.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On Tue, 5/13/2025 5:46 AM, Frank Slootweg wrote:

NY <me@privacy.net> wrote:

My PC takes ages (up to 10 minutes) to boot from "cold" (measuring until
all the background system tray applets have started and disk-thrashing
ceases, so most days I tend to put it to sleep rather than shutting it down. >>
I've noticed recently that it often (but not always) fails to start up
from sleep mode if a USB hard disk is connected as I press the power
soft-start switch. The hard disk light comes on continuously, as if the
memory contents are being restored from the HDD copy that sleep-mode
uses, but the graphics card does not turn on (no logon screen or
anything else displayed) and after a few seconds the power light reverts
to its orange "asleep" state. It's as if it wakes up momentarily and
then almost immediately goes back to sleep.

Are you sure it's sleeping? What you describe is hibernation.

When sleeping, you don't have to use "the power soft-start switch",
but can just wake up the computer with a keyboard press or 'mouse'
movement.

Also when sleeping, there is no situation of "memory contents are
being restored from the HDD copy that sleep-mode uses", because memory contents are preserved in RAM when sleeping. (Yes, there is some mode
when memory contents can also be copied to HDD, just in case the system
loses power when sleeping. I forgot what that mode is called and how
it's set. 'Hybrid Sleep'? (If so, I don't have Hybrid Sleep, because I
have a laptop, which doesn't need that mode.))

If the system is really sleeping (a normal sleep), it might be worth
your while to set/tick the 'Allow the computer to turn off this device
to save power' setting on the 'Power Management' tab of the (correct)
'USB Mass Strage Device' in the 'Universal Seerial Bus Controllers' tree
of Device Manager. Maybe with that setting, Windows might not touch the
USB HDD till the rest of the system has been woken up (sufficiently).

Is this likely to be a Windows issue or a BIOS/UEFI issue?

AFAIK, when sleeping, the BIOS/UEFI is not involved, because Windows
is still in charge [1]. But Paul would be better informed in this area.

I've not yet established whether the problem is confined to one specific
external HDD or whether other disks cause the same symptoms.

In any case, always use the *same* USB port, because the 'Allow the computer to turn off...' setting is per-port.

It's not a problem: I just need to remember not to plug the backup drive
in until I see a logo screen to prove that the PC is awake.

[1] If Windows isn't in control, it couldn't be configured (for example
with powercfg) to set which devices can and can not wake up the system.

The Sleep could be Hybrid Sleep, which upon a power failure
results in the reading of the Hiberfil.sys in order to restore
the session.

Paul

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

Paul <nospam@needed.invalid> wrote:

On Tue, 5/13/2025 5:46 AM, Frank Slootweg wrote:

NY <me@privacy.net> wrote:

My PC takes ages (up to 10 minutes) to boot from "cold" (measuring until >> all the background system tray applets have started and disk-thrashing
ceases, so most days I tend to put it to sleep rather than shutting it down.

I've noticed recently that it often (but not always) fails to start up
from sleep mode if a USB hard disk is connected as I press the power
soft-start switch. The hard disk light comes on continuously, as if the
memory contents are being restored from the HDD copy that sleep-mode
uses, but the graphics card does not turn on (no logon screen or
anything else displayed) and after a few seconds the power light reverts >> to its orange "asleep" state. It's as if it wakes up momentarily and
then almost immediately goes back to sleep.

Are you sure it's sleeping? What you describe is hibernation.

When sleeping, you don't have to use "the power soft-start switch",
but can just wake up the computer with a keyboard press or 'mouse' movement.

Also when sleeping, there is no situation of "memory contents are
being restored from the HDD copy that sleep-mode uses", because memory contents are preserved in RAM when sleeping. (Yes, there is some mode
when memory contents can also be copied to HDD, just in case the system loses power when sleeping. I forgot what that mode is called and how
it's set. 'Hybrid Sleep'? (If so, I don't have Hybrid Sleep, because I
have a laptop, which doesn't need that mode.))

If the system is really sleeping (a normal sleep), it might be worth
your while to set/tick the 'Allow the computer to turn off this device
to save power' setting on the 'Power Management' tab of the (correct)
'USB Mass Strage Device' in the 'Universal Seerial Bus Controllers' tree
of Device Manager. Maybe with that setting, Windows might not touch the
USB HDD till the rest of the system has been woken up (sufficiently).

Is this likely to be a Windows issue or a BIOS/UEFI issue?

AFAIK, when sleeping, the BIOS/UEFI is not involved, because Windows
is still in charge [1]. But Paul would be better informed in this area.

I've not yet established whether the problem is confined to one specific >> external HDD or whether other disks cause the same symptoms.

In any case, always use the *same* USB port, because the 'Allow the computer to turn off...' setting is per-port.

It's not a problem: I just need to remember not to plug the backup drive >> in until I see a logo screen to prove that the PC is awake.

[1] If Windows isn't in control, it couldn't be configured (for example with powercfg) to set which devices can and can not wake up the system.

The Sleep could be Hybrid Sleep, which upon a power failure
results in the reading of the Hiberfil.sys in order to restore
the session.

Yes, that's what I was/am thinking. But then, as I wrote, 'NY' doesn't
have to use "the power soft-start switch" to wake the system, but just
can wake it up with a keyboard press or 'mouse' movement.

Of course (short) pressing "the power soft-start switch" will also
wake from sleep, so perhaps 'NY' is just doing things the 'hard'/uncommon
way.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 2025-05-14 13:09, Frank Slootweg wrote:

Paul <nospam@needed.invalid> wrote:

On Tue, 5/13/2025 5:46 AM, Frank Slootweg wrote:

NY <me@privacy.net> wrote:

My PC takes ages (up to 10 minutes) to boot from "cold" (measuring until >>>> all the background system tray applets have started and disk-thrashing >>>> ceases, so most days I tend to put it to sleep rather than shutting it down.

I've noticed recently that it often (but not always) fails to start up >>>> from sleep mode if a USB hard disk is connected as I press the power
soft-start switch. The hard disk light comes on continuously, as if the >>>> memory contents are being restored from the HDD copy that sleep-mode
uses, but the graphics card does not turn on (no logon screen or
anything else displayed) and after a few seconds the power light reverts >>>> to its orange "asleep" state. It's as if it wakes up momentarily and
then almost immediately goes back to sleep.

Are you sure it's sleeping? What you describe is hibernation.

When sleeping, you don't have to use "the power soft-start switch",
but can just wake up the computer with a keyboard press or 'mouse'
movement.

Also when sleeping, there is no situation of "memory contents are
being restored from the HDD copy that sleep-mode uses", because memory
contents are preserved in RAM when sleeping. (Yes, there is some mode
when memory contents can also be copied to HDD, just in case the system
loses power when sleeping. I forgot what that mode is called and how
it's set. 'Hybrid Sleep'? (If so, I don't have Hybrid Sleep, because I
have a laptop, which doesn't need that mode.))

If the system is really sleeping (a normal sleep), it might be worth
your while to set/tick the 'Allow the computer to turn off this device
to save power' setting on the 'Power Management' tab of the (correct)
'USB Mass Strage Device' in the 'Universal Seerial Bus Controllers' tree >>> of Device Manager. Maybe with that setting, Windows might not touch the
USB HDD till the rest of the system has been woken up (sufficiently).

Is this likely to be a Windows issue or a BIOS/UEFI issue?

AFAIK, when sleeping, the BIOS/UEFI is not involved, because Windows
is still in charge [1]. But Paul would be better informed in this area.

I've not yet established whether the problem is confined to one specific >>>> external HDD or whether other disks cause the same symptoms.

In any case, always use the *same* USB port, because the 'Allow the
computer to turn off...' setting is per-port.

It's not a problem: I just need to remember not to plug the backup drive >>>> in until I see a logo screen to prove that the PC is awake.

[1] If Windows isn't in control, it couldn't be configured (for example
with powercfg) to set which devices can and can not wake up the system.

The Sleep could be Hybrid Sleep, which upon a power failure
results in the reading of the Hiberfil.sys in order to restore
the session.

Yes, that's what I was/am thinking. But then, as I wrote, 'NY' doesn't have to use "the power soft-start switch" to wake the system, but just
can wake it up with a keyboard press or 'mouse' movement.

Of course (short) pressing "the power soft-start switch" will also
wake from sleep, so perhaps 'NY' is just doing things the 'hard'/uncommon way.

This is usually configurable in the BIOS, what will wakeup or not the
machine.

--
Cheers, Carlos.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

Carlos E.R. <robin_listas@es.invalid> wrote:

On 2025-05-14 13:09, Frank Slootweg wrote:

Paul <nospam@needed.invalid> wrote:

[...]

The Sleep could be Hybrid Sleep, which upon a power failure
results in the reading of the Hiberfil.sys in order to restore
the session.

Yes, that's what I was/am thinking. But then, as I wrote, 'NY' doesn't have to use "the power soft-start switch" to wake the system, but just
can wake it up with a keyboard press or 'mouse' movement.

Of course (short) pressing "the power soft-start switch" will also
wake from sleep, so perhaps 'NY' is just doing things the 'hard'/uncommon way.

This is usually configurable in the BIOS, what will wakeup or not the machine.

Yes, it's usually configurable in the BIOS, but it also is (or at
least was in earlier Windows (I used it in Vista to shut up a laptop
which was waking up all the time)) configurable with the 'powercfg /deviceenablewake' (and /devicedisablewake) command.

However just now, I can't get it to work, because 'powercfg
/devicequery wake_armed' returns 'NONE' and 'powercfg /deviceenablewake "Standard PS/2 Keyboard"' says 'You do not have permission to enable or
disable device wake.', while I'm Administrator.

Anyway, I can wake up (from sleep), my Windows 8.1, 10 and 11 laptops
with the keyboard and the 'mouse' (touchpad) (and of course also with
"the power soft-start switch" (which is the power button/key on the
keyboard; short-press is wakeup, long press is force power off)).

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 12/05/2025 23:14, NY wrote:

My PC takes ages (up to 10 minutes) to boot from "cold" (measuring until
all the background system tray applets have started and disk-thrashing ceases, so most days I tend to put it to sleep rather than shutting it
down.

I've noticed recently that it often (but not always) fails to start up
from sleep mode if a USB hard disk is connected as I press the power soft-start switch. The hard disk light comes on continuously, as if the memory contents are being restored from the HDD copy that sleep-mode
uses, but the graphics card does not turn on (no logon screen or
anything else displayed) and after a few seconds the power light reverts
to its orange "asleep" state. It's as if it wakes up momentarily and
then almost immediately goes back to sleep.

Is this likely to be a Windows issue or a BIOS/UEFI issue?

I've not yet established whether the problem is confined to one specific external HDD or whether other disks cause the same symptoms.

It's not a problem: I just need to remember not to plug the backup drive
in until I see a logo screen to prove that the PC is awake.

I've had either a boot failure or slow booting on PCs I'm working on.
The cause is often a USB stick still plugged in. Remove it & the PC
boots although the first boot may still be slow.
Sometimes that's all that's needed, other times the USB device needs
attention.


--
Regards
wasbit

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 2025-05-14 14:41, Frank Slootweg wrote:

Carlos E.R. <robin_listas@es.invalid> wrote:

On 2025-05-14 13:09, Frank Slootweg wrote:

Paul <nospam@needed.invalid> wrote:

[...]

The Sleep could be Hybrid Sleep, which upon a power failure
results in the reading of the Hiberfil.sys in order to restore
the session.

Yes, that's what I was/am thinking. But then, as I wrote, 'NY' doesn't >>> have to use "the power soft-start switch" to wake the system, but just
can wake it up with a keyboard press or 'mouse' movement.

Of course (short) pressing "the power soft-start switch" will also
wake from sleep, so perhaps 'NY' is just doing things the 'hard'/uncommon >>> way.

This is usually configurable in the BIOS, what will wakeup or not the
machine.

Yes, it's usually configurable in the BIOS, but it also is (or at
least was in earlier Windows (I used it in Vista to shut up a laptop
which was waking up all the time)) configurable with the 'powercfg /deviceenablewake' (and /devicedisablewake) command.

However just now, I can't get it to work, because 'powercfg
/devicequery wake_armed' returns 'NONE' and 'powercfg /deviceenablewake "Standard PS/2 Keyboard"' says 'You do not have permission to enable or disable device wake.', while I'm Administrator.

Argh. :-(

Anyway, I can wake up (from sleep), my Windows 8.1, 10 and 11 laptops
with the keyboard and the 'mouse' (touchpad) (and of course also with
"the power soft-start switch" (which is the power button/key on the
keyboard; short-press is wakeup, long press is force power off)).

I don't have a system running Windows this instant (it is virtual), but
my Linux machines do not wake up touching the mouse; I would find that inconvenient. Think of pushing the mouse away in the table to place the
cup of tea or a book :-)

--
Cheers, Carlos.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

Carlos E.R. <robin_listas@es.invalid> wrote:

On 2025-05-14 14:41, Frank Slootweg wrote:

Carlos E.R. <robin_listas@es.invalid> wrote:

On 2025-05-14 13:09, Frank Slootweg wrote:

Paul <nospam@needed.invalid> wrote:

[...]

The Sleep could be Hybrid Sleep, which upon a power failure
results in the reading of the Hiberfil.sys in order to restore
the session.

Yes, that's what I was/am thinking. But then, as I wrote, 'NY' doesn't
have to use "the power soft-start switch" to wake the system, but just >>> can wake it up with a keyboard press or 'mouse' movement.

Of course (short) pressing "the power soft-start switch" will also >>> wake from sleep, so perhaps 'NY' is just doing things the 'hard'/uncommon >>> way.

This is usually configurable in the BIOS, what will wakeup or not the
machine.

Yes, it's usually configurable in the BIOS, but it also is (or at
least was in earlier Windows (I used it in Vista to shut up a laptop
which was waking up all the time)) configurable with the 'powercfg /deviceenablewake' (and /devicedisablewake) command.

However just now, I can't get it to work, because 'powercfg
/devicequery wake_armed' returns 'NONE' and 'powercfg /deviceenablewake "Standard PS/2 Keyboard"' says 'You do not have permission to enable or disable device wake.', while I'm Administrator.

Argh. :-(

Anyway, I can wake up (from sleep), my Windows 8.1, 10 and 11 laptops with the keyboard and the 'mouse' (touchpad) (and of course also with
"the power soft-start switch" (which is the power button/key on the keyboard; short-press is wakeup, long press is force power off)).

I don't have a system running Windows this instant (it is virtual), but
my Linux machines do not wake up touching the mouse; I would find that inconvenient. Think of pushing the mouse away in the table to place the
cup of tea or a book :-)

Yes, that can be inconvenient/annoying.

I have a laptop with only it's touchpad, so no problem.

My wife has a (sort of) 'desktop' with a wireless mouse, which she
actually uses to wake up the system. The mouse is on a mouse-pad, so we
don't have the problems you mention , but yes, sometimes it will wake up unintentionally. That's why I wish the 'powercfg'stuff would still work,
in case we need it.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 15/05/2025 16:28, Frank Slootweg wrote:

I don't have a system running Windows this instant (it is virtual), but
my Linux machines do not wake up touching the mouse; I would find that
inconvenient. Think of pushing the mouse away in the table to place the
cup of tea or a book :-)

Yes, that can be inconvenient/annoying.

I remember in the early days of Sun Workstations in the mid 1980s,
hearing of a problem where the Sun would never go into screen-saver
mode, even if the mouse and keyboard were not touched for the timeout
period. It was eventually traced to a "newfangled" optical mouse (in the
days when they need a special metal mouse pad with ruled lines of one
colour for vertical and another for horizontal) which was moving very
slightly as people walked across the floor of the room and so was
constantly detecting mouse movement ;-)

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 14/05/2025 12:09, Frank Slootweg wrote:

Paul <nospam@needed.invalid> wrote:

The Sleep could be Hybrid Sleep, which upon a power failure
results in the reading of the Hiberfil.sys in order to restore
the session.

Yes, that's what I was/am thinking. But then, as I wrote, 'NY' doesn't have to use "the power soft-start switch" to wake the system, but just
can wake it up with a keyboard press or 'mouse' movement.

Of course (short) pressing "the power soft-start switch" will also
wake from sleep, so perhaps 'NY' is just doing things the 'hard'/uncommon way.

The PC doesn't wake on keyboard or mouse movement, and only resumes when
the power sort switch is pressed. When I put it to "sleep" (as Windows
calls it) there is period of about 10 seconds while it seems to dump the
RAM to HDD but when I resume it (by pressing power button) the logon
screen appears too quickly to have read from HDD.

So I think it is using RAM to resume, but sometimes reading from HDD if necessary. And always saving to HDD in case this is needed.


I wonder if it could be contact bounce in the power soft switch. You've
think the hardware would have a fairly long timeout to debounce the
switch, to avoid phantom on-off events, but maybe a capacitor is not
imposing a long enough delay in the debounce logic if the external HDD
is drawing current.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 13/05/2025 10:46, Frank Slootweg wrote:

NY <me@privacy.net> wrote:

My PC takes ages (up to 10 minutes) to boot from "cold" (measuring until
all the background system tray applets have started and disk-thrashing
ceases, so most days I tend to put it to sleep rather than shutting it down. >>
I've noticed recently that it often (but not always) fails to start up
from sleep mode if a USB hard disk is connected as I press the power
soft-start switch. The hard disk light comes on continuously, as if the
memory contents are being restored from the HDD copy that sleep-mode
uses, but the graphics card does not turn on (no logon screen or
anything else displayed) and after a few seconds the power light reverts
to its orange "asleep" state. It's as if it wakes up momentarily and
then almost immediately goes back to sleep.

Are you sure it's sleeping? What you describe is hibernation.

When sleeping, you don't have to use "the power soft-start switch",
but can just wake up the computer with a keyboard press or 'mouse'
movement.

Also when sleeping, there is no situation of "memory contents are
being restored from the HDD copy that sleep-mode uses", because memory contents are preserved in RAM when sleeping. (Yes, there is some mode
when memory contents can also be copied to HDD, just in case the system
loses power when sleeping. I forgot what that mode is called and how
it's set. 'Hybrid Sleep'? (If so, I don't have Hybrid Sleep, because I
have a laptop, which doesn't need that mode.))

You're right. It is hibernating - when I put it into that state, there
is a period of frantic disk activity as it is dumping the RAM to HDD,
and then the PSU switches off.

It's described on the Start Menu | Power menu as "Sleep" but it's not ;-)

Someone asked whether the external HDD is drawing current from the +5V
standby line and preventing the PSU "noticing" that I have pressed the
button. It's not quite that, because the PSU starts up (power light on
front panel comes on, HDDs whirr) but the graphics card never starts
sending a signal and after a few seconds of disk activity the power
light goes back to its "I'm asleep" state (changes from white to amber
LED on front panel).


Occasionally, with or without an external HDD connected, the PC has
displayed its BIOS screen and then a "Resuming Windows" banner. Normal
service is eventually restored after a longer-than-normal period of
faffing around during which time Windows isn't fully responsive to
starting apps from icons, dragging windows, opening the Start Menu. But
even that is preferable to a 10-minute wait for a cold power-off boot
from shutdown.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

NY <me@privacy.net> wrote:

On 14/05/2025 12:09, Frank Slootweg wrote:

Paul <nospam@needed.invalid> wrote:

The Sleep could be Hybrid Sleep, which upon a power failure
results in the reading of the Hiberfil.sys in order to restore
the session.

Yes, that's what I was/am thinking. But then, as I wrote, 'NY' doesn't have to use "the power soft-start switch" to wake the system, but just
can wake it up with a keyboard press or 'mouse' movement.

Of course (short) pressing "the power soft-start switch" will also
wake from sleep, so perhaps 'NY' is just doing things the 'hard'/uncommon way.

The PC doesn't wake on keyboard or mouse movement, and only resumes when
the power sort switch is pressed.

When (really) sleeping, the PC should - or at least should be able to
be configured to - wake on keyboard or mouse movement.

When I put it to "sleep" (as Windows
calls it) there is period of about 10 seconds while it seems to dump the
RAM to HDD but when I resume it (by pressing power button) the logon
screen appears too quickly to have read from HDD.

As (also I) mentioned elsewhere in the thread, your PC is probably
using 'Hybrid Sleep', which has the behaviour you describe.

As I mentioned, I have a laptop which doesn't need/have Hybdrid Sleep.

Probably others, who have a 'desktop' like you, can point out where
(probably in the 'Power Options' applet), you can check/change the
Hybdrid Sleep setting.

So I think it is using RAM to resume, but sometimes reading from HDD if necessary. And always saving to HDD in case this is needed.

Yes, it;s using RAM if that's still valid (i.e. didn't lose power) and
will resume from HDD when RAM content is lost.

I wonder if it could be contact bounce in the power soft switch. You've
think the hardware would have a fairly long timeout to debounce the
switch, to avoid phantom on-off events, but maybe a capacitor is not
imposing a long enough delay in the debounce logic if the external HDD
is drawing current.

Try to get the keyboard/mouse wake the PC from sleep. That's the
normal way and you then don't have to use the power soft switch. (I
can't help will that, because I don't have a 'desktop'.)

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

NY <me@privacy.net> wrote:

On 13/05/2025 10:46, Frank Slootweg wrote:

NY <me@privacy.net> wrote:

My PC takes ages (up to 10 minutes) to boot from "cold" (measuring until >> all the background system tray applets have started and disk-thrashing
ceases, so most days I tend to put it to sleep rather than shutting it down.

I've noticed recently that it often (but not always) fails to start up
from sleep mode if a USB hard disk is connected as I press the power
soft-start switch. The hard disk light comes on continuously, as if the
memory contents are being restored from the HDD copy that sleep-mode
uses, but the graphics card does not turn on (no logon screen or
anything else displayed) and after a few seconds the power light reverts >> to its orange "asleep" state. It's as if it wakes up momentarily and
then almost immediately goes back to sleep.

Are you sure it's sleeping? What you describe is hibernation.

When sleeping, you don't have to use "the power soft-start switch",
but can just wake up the computer with a keyboard press or 'mouse' movement.

Also when sleeping, there is no situation of "memory contents are
being restored from the HDD copy that sleep-mode uses", because memory contents are preserved in RAM when sleeping. (Yes, there is some mode
when memory contents can also be copied to HDD, just in case the system loses power when sleeping. I forgot what that mode is called and how
it's set. 'Hybrid Sleep'? (If so, I don't have Hybrid Sleep, because I
have a laptop, which doesn't need that mode.))

You're right. It is hibernating - when I put it into that state, there
is a period of frantic disk activity as it is dumping the RAM to HDD,
and then the PSU switches off.

As I mentioned, I have a laptop, not a 'desktop', but AFAIK, when
putting the system to sleep, the PSU should not switch off completely,
but should keep the RAM, etc., etc. alive, but should switch off disks,
fans, etc..

Paul can probably help with that (or set me straight).

It's described on the Start Menu | Power menu as "Sleep" but it's not ;-)

Someone asked whether the external HDD is drawing current from the +5V standby line and preventing the PSU "noticing" that I have pressed the button. It's not quite that, because the PSU starts up (power light on
front panel comes on, HDDs whirr) but the graphics card never starts
sending a signal and after a few seconds of disk activity the power
light goes back to its "I'm asleep" state (changes from white to amber
LED on front panel).

Occasionally, with or without an external HDD connected, the PC has
displayed its BIOS screen and then a "Resuming Windows" banner. Normal service is eventually restored after a longer-than-normal period of
faffing around during which time Windows isn't fully responsive to
starting apps from icons, dragging windows, opening the Start Menu. But
even that is preferable to a 10-minute wait for a cold power-off boot
from shutdown.

When the system says "Resuming Windows", it's resuming from
hibernation (i.e. a complete copy of the RAM content has be saved/
restored to/from disk) and the system resumes from where it left off,
with all your active programs intact.

I don't think - but could be wrong - it gives that message when it
does a "Fast startup", i.e. it only loads Windows itself from a copy on
HDD and you still need to log in and all your programs still have to get
or be started.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On Fri, 5/16/2025 10:35 AM, NY wrote:

On 14/05/2025 12:09, Frank Slootweg wrote:

Paul <nospam@needed.invalid> wrote:

The Sleep could be Hybrid Sleep, which upon a power failure
results in the reading of the Hiberfil.sys in order to restore
the session.

   Yes, that's what I was/am thinking. But then, as I wrote, 'NY' doesn't >> have to use "the power soft-start switch" to wake the system, but just
can wake it up with a keyboard press or 'mouse' movement.

   Of course (short) pressing "the power soft-start switch" will also
wake from sleep, so perhaps 'NY' is just doing things the 'hard'/uncommon
way.

The PC doesn't wake on keyboard or mouse movement, and only resumes when the power sort switch is pressed. When I put it to "sleep" (as Windows calls it) there is period of about 10 seconds while it seems to dump the RAM to HDD but when I resume it (by

pressing power button) the logon screen appears too quickly to have read from HDD.

So I think it is using RAM to resume, but sometimes reading from HDD if necessary. And always saving to HDD in case this is needed.

I wonder if it could be contact bounce in the power soft switch. You've think the hardware would have a fairly long timeout to debounce the switch, to avoid phantom on-off events, but maybe a capacitor is not imposing a long enough delay in the

debounce logic if the external HDD is drawing current.

Of course equipment has debounce where it is needed :-)
Do you think we are primitive cavemen ? (We are actually.)

https://www.digikey.com/en/articles/how-to-implement-hardware-debounce-for-switches-and-relays

As hobbyists, we could use LS279 for cleaning up our SPDT
switch inputs. PC buttons on the other hand, are SPST momentary contact.

https://www.circuits-diy.com/wp-content/uploads/2021/01/74ls279-pinout.jpg

You know the Soft Power switch on the front of the PC, *does not*
control the power supply directly. The switch signal goes
through either the SuperIO or the Southbridge, before the
signal becomes PSON- and drives the PSU wire on the 24 pin main
connector. There is some amount of conditioning for that signal
as a result.

The ATX PSU itself, does not have a logic chip connected to PSON-.
(Any digital behaviors are pure coincidence. It's not a TTL
logic board by any stretch.) A transistor is used to "parse" that signal.

The drive current to operate that logic-like signal, is usually
pretty low, like 2mA loading might be suspected. But the motherboard
designers have had other suspicions. I've seen designs where the
motherboard is ready for any current up to 64mA on that signal :-)

There seems to be some "tension", between what power supply
designers hang off that signal, and what motherboard designers
drive with. Some motherboard designers have used 8mA drive, and then
the driver signal mysteriously fails to work any more. This is
one of the reasons, that the odd motherboard designer got
angry about the bad dining habits of PSU makers and fitted
an output driver with "hair on its knuckles". When you use
chips with 64mA drive, smoke comes out of things, on a failure :-)
That's the best part, of using muscular parts and getting
them to fight one another, is the plastic package cracking
and smoke comes out.

We use web pages like this for inspiration. PSON- is on the left,
in the vertical middle of the circuit. R23 further over, is a
4.7K pullup. That means, when the motherboard asserts PSON- by
driving to ground, it has to sink ~1 mA of current from R23. Hardly
seems a reason to have a 64ma drive 74F07 for the job :-)
But if your motherboards have significant field returns with
a blown PSON- driving circuit as the reason for failure, there
is an "I will fix your wagon" response from the motherboard side.

http://www.pavouk.org/hw/en_atxps.html

( https://assets.nexperia.com/documents/data-sheet/74F07.pdf ) 64mA drive, logic cct

The most current you can get in a logic chip, is on the
order of 2 amperes on a specific CMOS chip. It's not an
ordinary logic gate, because if you feed in 5V on the
logic input, 4.8V comes out of the output. If you run
the 4.8V signal into another one of those, 4.6V comes out.
It is, in essence, an analog amplifier with a voltage gain of
less than 1, so if the circuits are cascaded enough
times, you might eventually fail to make a logic level.
But if you're a motherboard designer, who is tired of
some brand of PSU damaging your motherboards, you can
"fix their wagon" with a 2 ampere driving circuit.

I don't see a number specified here, for the pullup current flow.
As far as I know, the "natural assumption" is no more than
one Schottky load or 2mA should be sufficient, and an 8mA drive
motherboard should be able to safely turn it on and off.

ATX12V_PSDG_2_2_public_br2.pdf

https://web.archive.org/web/20070112183223if_/http://www.formfactors.org:80/developer/specs/ATX12V_PSDG_2_2_public_br2.pdf

Paul

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On Fri, 5/16/2025 10:28 AM, NY wrote:

On 13/05/2025 10:46, Frank Slootweg wrote:

NY <me@privacy.net> wrote:

My PC takes ages (up to 10 minutes) to boot from "cold" (measuring until >>> all the background system tray applets have started and disk-thrashing
ceases, so most days I tend to put it to sleep rather than shutting it down.

I've noticed recently that it often (but not always) fails to start up
from sleep mode if a USB hard disk is connected as I press the power
soft-start switch. The hard disk light comes on continuously, as if the
memory contents are being restored from the HDD copy that sleep-mode
uses, but the graphics card does not turn on (no logon screen or
anything else displayed) and after a few seconds the power light reverts >>> to its orange "asleep" state. It's as if it wakes up momentarily and
then almost immediately goes back to sleep.

   Are you sure it's sleeping? What you describe is hibernation.

   When sleeping, you don't have to use "the power soft-start switch",
but can just wake up the computer with a keyboard press or 'mouse'
movement.

   Also when sleeping, there is no situation of "memory contents are
being restored from the HDD copy that sleep-mode uses", because memory
contents are preserved in RAM when sleeping. (Yes, there is some mode
when memory contents can also be copied to HDD, just in case the system
loses power when sleeping. I forgot what that mode is called and how
it's set. 'Hybrid Sleep'? (If so, I don't have Hybrid Sleep, because I
have a laptop, which doesn't need that mode.))

You're right. It is hibernating - when I put it into that state, there is a period of frantic disk activity as it is dumping the RAM to HDD, and then the PSU switches off.

It's described on the Start Menu | Power menu as "Sleep" but it's not ;-)

Someone asked whether the external HDD is drawing current from the +5V standby line and preventing the PSU "noticing" that I have pressed the button. It's not quite that, because the PSU starts up (power light on front panel comes on, HDDs whirr) but

the graphics card never starts sending a signal and after a few seconds of disk activity the power light goes back to its "I'm asleep" state (changes from white to amber LED on front panel).

Occasionally, with or without an external HDD connected, the PC has displayed its BIOS screen and then a "Resuming Windows" banner. Normal service is eventually restored after a longer-than-normal period of faffing around during which time Windows isn'

t fully responsive to starting apps from icons, dragging windows, opening the Start Menu. But even that is preferable to a 10-minute wait for a cold power-off boot from shutdown.

If you open Settings and type "Relia" and bring up the Reliability Monitor, what kind of events do you see logged in there ?

There is no particular reason why definitive results will be logged.
I have a "crashing issue" where the only log entry is "Windows was
uncleanly shut down" or equivalent, as evidence something naughty
is going on.

The "powercfg" command, has various logging options for things
related to state changes. It can, for example, tentatively identify ("LastWake") what woke the computer, like Power Button.

If you use Device Manager, and untick the box on mouse and keyboard,
to disable the "Allow this device to wake the computer", then you can
make it so that only the Power Button is the single ACPI event capable
of waking the machine. And Powercfg should record in LastWake, that
it was the Power Button that woke it up.

The PC should also be awakened by the RTC timer. Microsoft has
used that on occasion, to "do things when you are not present",
but they don't seem to be using that today. Third party programs
can still fiddle with your PC when you are not around, and this
can lead to confusion about exactly what state the machine
should be in. There have been cases in the past, where people get
up in the morning, their PC is running... and they did not leave
it that way the day before. And this is an RTC based waking event.

The RTC has limited waking range, and I don't think it can go past
31 days or so. To set a wake time of seven months from now,
requires the software to reset the timer for another 30/31 days,
and do that, seven times.

But as long as OSes do not go back to sleep properly
after a Wake Event like that, there will be lots of possibilities
for confusion.

You could point a web cam connected to a second machine, at
the suspicious machine, and check the fossil record later and
see if anything is happening when you are not there.

*******

It should not take ten minutes for a machine to cold boot :-)
Even my Optiplex 780 with E8400 in it, doesn't take that
long to boot Windows 10 22H2. By sticking an HD6450 video card
in it, I convinced it to run 22H2 and it still functions.

Part of the boot process, is Windows Defender scanning "critical areas"
at startup. If an HDD is being used, the slow seek of HDD mean
there will be a delay until that scan is finished. But I bet with
a little effort, you could suppress that scan. Perhaps declaring the
entire C: drive as an "Exclusion Folder" would do it :-) <snicker>
as appropriate.

Sysinternals Process Monitor can do a boot trace. That's if you
wanted to study what is going on at startup. The trace is not WinDbg
quality, but it does track what files are being opened for example.
Even if you replace the HDD with an SSD, if you don't have a
particularly powerful CPU, it can still take a while to finish a
scan at startup.

https://learn.microsoft.com/en-us/sysinternals/downloads/procmon

My laptop has a single core processor at 2.2GHz and it
does not need ten minutes to start. it runs Windows 10.
I don't plan on bodging in Win11 on the laptop,
as the RAM situation is pretty bad on the machine.

Paul

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 2025-05-16 16:35, NY wrote:

On 14/05/2025 12:09, Frank Slootweg wrote:

The PC doesn't wake on keyboard or mouse movement, and only resumes when
the power sort switch is pressed. When I put it to "sleep" (as Windows
calls it) there is period of about 10 seconds while it seems to dump the
RAM to HDD but when I resume it (by pressing power button) the logon
screen appears too quickly to have read from HDD.

Hybrid mode. If the battery runs out, it can recover from disk when powered.

So I think it is using RAM to resume, but sometimes reading from HDD if necessary. And always saving to HDD in case this is needed.

I wonder if it could be contact bounce in the power soft switch. You've
think the hardware would have a fairly long timeout to debounce the
switch, to avoid phantom on-off events, but maybe a capacitor is not
imposing a long enough delay in the debounce logic if the external HDD
is drawing current.

Computers normally debounce in software.

--
Cheers,
Carlos E.R.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

Carlos E. R. <robin_listas@es.invalid> wrote:

On 2025-05-16 16:35, NY wrote:

On 14/05/2025 12:09, Frank Slootweg wrote:

The PC doesn't wake on keyboard or mouse movement, and only resumes when the power sort switch is pressed. When I put it to "sleep" (as Windows calls it) there is period of about 10 seconds while it seems to dump the RAM to HDD but when I resume it (by pressing power button) the logon
screen appears too quickly to have read from HDD.

Hybrid mode. If the battery runs out, it can recover from disk when powered.

Hybrid Sleep. But I'm sure the OP ('NY') has a 'desktop', not a
laptop, i.e. no battery.

Actually these days, 'desktop's have Hybrid Sleep, but laptops do not.

Why? Because laptops don't need it any more. 'Sleep' is now 'Modern
Standby', i.e. not fully sleeping and 'Hibernate' is now 'Adaptive
hibernate'. 'Adaptive hibernate' will make sure that the system does not
lose too much (percent) battery capacity when sleeping and when it does,
it will do a full hibernate. So for these laptops, 'Hybrid Sleep'
doesn't offer any additional features.

So I think it is using RAM to resume, but sometimes reading from HDD if necessary. And always saving to HDD in case this is needed.

[...]

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 2025-05-17 19:51, Frank Slootweg wrote:

Carlos E. R. <robin_listas@es.invalid> wrote:

On 2025-05-16 16:35, NY wrote:

On 14/05/2025 12:09, Frank Slootweg wrote:

The PC doesn't wake on keyboard or mouse movement, and only resumes when >>> the power sort switch is pressed. When I put it to "sleep" (as Windows
calls it) there is period of about 10 seconds while it seems to dump the >>> RAM to HDD but when I resume it (by pressing power button) the logon
screen appears too quickly to have read from HDD.

Hybrid mode. If the battery runs out, it can recover from disk when powered.

Hybrid Sleep. But I'm sure the OP ('NY') has a 'desktop', not a
laptop, i.e. no battery.

Still, they support sleep, using the PSU.

Actually these days, 'desktop's have Hybrid Sleep, but laptops do not.

AFAIK, mine does (Linux, though).

Why? Because laptops don't need it any more. 'Sleep' is now 'Modern Standby', i.e. not fully sleeping and 'Hibernate' is now 'Adaptive hibernate'. 'Adaptive hibernate' will make sure that the system does not
lose too much (percent) battery capacity when sleeping and when it does,
it will do a full hibernate. So for these laptops, 'Hybrid Sleep'
doesn't offer any additional features.

So I think it is using RAM to resume, but sometimes reading from HDD if
necessary. And always saving to HDD in case this is needed.

[...]

--
Cheers,
Carlos E.R.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On Sat, 5/17/2025 1:51 PM, Frank Slootweg wrote:

Carlos E. R. <robin_listas@es.invalid> wrote:

On 2025-05-16 16:35, NY wrote:

On 14/05/2025 12:09, Frank Slootweg wrote:

The PC doesn't wake on keyboard or mouse movement, and only resumes when >>> the power sort switch is pressed. When I put it to "sleep" (as Windows
calls it) there is period of about 10 seconds while it seems to dump the >>> RAM to HDD but when I resume it (by pressing power button) the logon
screen appears too quickly to have read from HDD.

Hybrid mode. If the battery runs out, it can recover from disk when powered.

Hybrid Sleep. But I'm sure the OP ('NY') has a 'desktop', not a
laptop, i.e. no battery.

Actually these days, 'desktop's have Hybrid Sleep, but laptops do not.

Why? Because laptops don't need it any more. 'Sleep' is now 'Modern Standby', i.e. not fully sleeping and 'Hibernate' is now 'Adaptive hibernate'. 'Adaptive hibernate' will make sure that the system does not
lose too much (percent) battery capacity when sleeping and when it does,
it will do a full hibernate. So for these laptops, 'Hybrid Sleep'
doesn't offer any additional features.

So I think it is using RAM to resume, but sometimes reading from HDD if
necessary. And always saving to HDD in case this is needed.

For the OPS case, if you turned off Hibernation,
would this clarify whether sleep recovery was working ?
Since it cannot involve the hiberfile, if there is no hiberfile.

powercfg /h off

It's still OK for an OS to poke a hard drive, during
driver warm starts coming out of sleep. We would not expect
a USB HDD to be entirely ignored coming out of sleep.
Restoring the "HDD state" is all part of restoring the
runtime environment in a thorough manner.

Paul

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

Carlos E. R. <robin_listas@es.invalid> wrote:

On 2025-05-17 19:51, Frank Slootweg wrote:

Carlos E. R. <robin_listas@es.invalid> wrote:

On 2025-05-16 16:35, NY wrote:

On 14/05/2025 12:09, Frank Slootweg wrote:

The PC doesn't wake on keyboard or mouse movement, and only resumes when >>> the power sort switch is pressed. When I put it to "sleep" (as Windows >>> calls it) there is period of about 10 seconds while it seems to dump the >>> RAM to HDD but when I resume it (by pressing power button) the logon
screen appears too quickly to have read from HDD.

Hybrid mode. If the battery runs out, it can recover from disk when
powered.

Hybrid Sleep. But I'm sure the OP ('NY') has a 'desktop', not a
laptop, i.e. no battery.

Still, they support sleep, using the PSU.

Yes, as they should. Somewhere in this thread there's a suggestion
that a sleep of a 'desktop' shuts down all power. That's of course not
true, but as I don't have a desktop, I can't confirm.

Actually these days, 'desktop's have Hybrid Sleep, but laptops do not.

AFAIK, mine does (Linux, though).

How 'old'/new is that laptop? AFAIK, it's mainly laptops which came
with Windows 11 and perhaps some late Windows 10 laptops. I got my
Windows 11 laptop in July 2022.

Why? Because laptops don't need it any more. 'Sleep' is now 'Modern Standby', i.e. not fully sleeping and 'Hibernate' is now 'Adaptive hibernate'. 'Adaptive hibernate' will make sure that the system does not lose too much (percent) battery capacity when sleeping and when it does,
it will do a full hibernate. So for these laptops, 'Hybrid Sleep'
doesn't offer any additional features.

[...]

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 2025-05-18 11:32, Frank Slootweg wrote:

Carlos E. R. <robin_listas@es.invalid> wrote:

On 2025-05-17 19:51, Frank Slootweg wrote:

Carlos E. R. <robin_listas@es.invalid> wrote:

On 2025-05-16 16:35, NY wrote:

On 14/05/2025 12:09, Frank Slootweg wrote:

The PC doesn't wake on keyboard or mouse movement, and only resumes when >>>>> the power sort switch is pressed. When I put it to "sleep" (as Windows >>>>> calls it) there is period of about 10 seconds while it seems to dump the >>>>> RAM to HDD but when I resume it (by pressing power button) the logon >>>>> screen appears too quickly to have read from HDD.

Hybrid mode. If the battery runs out, it can recover from disk when
powered.

Hybrid Sleep. But I'm sure the OP ('NY') has a 'desktop', not a
laptop, i.e. no battery.

Still, they support sleep, using the PSU.

Yes, as they should. Somewhere in this thread there's a suggestion
that a sleep of a 'desktop' shuts down all power. That's of course not
true, but as I don't have a desktop, I can't confirm.

It needs power, at least on the RAM and some chips to detect a keyboard
press and activate wakeup. I do not know if CPU is unpowered.

Actually these days, 'desktop's have Hybrid Sleep, but laptops do not. >>

AFAIK, mine does (Linux, though).

How 'old'/new is that laptop? AFAIK, it's mainly laptops which came
with Windows 11 and perhaps some late Windows 10 laptops. I got my
Windows 11 laptop in July 2022.

Two or three years old. Lenovo.


--
Cheers,
Carlos E.R.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On Sun, 5/18/2025 5:58 AM, Carlos E. R. wrote:

On 2025-05-18 11:32, Frank Slootweg wrote:

Carlos E. R. <robin_listas@es.invalid> wrote:

On 2025-05-17 19:51, Frank Slootweg wrote:

Carlos E. R. <robin_listas@es.invalid> wrote:

On 2025-05-16 16:35, NY wrote:

On 14/05/2025 12:09, Frank Slootweg wrote:

The PC doesn't wake on keyboard or mouse movement, and only resumes when >>>>>> the power sort switch is pressed. When I put it to "sleep" (as Windows >>>>>> calls it) there is period of about 10 seconds while it seems to dump the >>>>>> RAM to HDD but when I resume it (by pressing power button) the logon >>>>>> screen appears too quickly to have read from HDD.

Hybrid mode. If the battery runs out, it can recover from disk when
powered.

    Hybrid Sleep. But I'm sure the OP ('NY') has a 'desktop', not a >>>> laptop, i.e. no battery.

Still, they support sleep, using the PSU.

   Yes, as they should. Somewhere in this thread there's a suggestion
that a sleep of a 'desktop' shuts down all power. That's of course not
true, but as I don't have a desktop, I can't confirm.

It needs power, at least on the RAM and some chips to detect a
keyboard press and activate wakeup. I do not know if CPU is unpowered.

You would need Mechanical Off to remove all power.

Otherwise, if the PSU is switched on at the back, the +5VSB and
related circuits provide supervisory power.

To make a functional USB wake event, VBUS must be powered by +5VSB
so that the keyboard or mouse has +5VSB on it. The logic circuits
on board, some of those run off 3VSB or +5VSB. Things like PME
(Power Management Event) signal, those enter the logic as an indication
service is required.

During Sleep, DRAM runs off +5VSB. The DRAM are in "AutoRefresh",
a kind of instruction useful for parking them. The DRAM have a
signal from the clockgen, so the chip is clocked. The voltage to
the DRAM can be slightly lower. (The voltage to the DRAM typically
comes from a small switching converter, powered by +5VSB). This
means the DRAM has to do a logical-OR of power sources. The DRAM can
also be a major drain (4W or 7W on a desktop, not a small amount
when the max output is 15W from the ATX supply on +5VSB).

If you have Wake On LAN enabled, the Ethernet chip may receive
core power so it can "listen". The receivers on the PHY likely
take a lot less power, than transmitting packets with 50 ohm drive
to the Ethernet isolation transformers in the black epoxy pack.
Switching off the Ethernet (because most people are not set up
for WOL Wake On LAN) is best but might only save you a watt.

In Hibernate, the converter for the DRAM does not have to be
enabled and can be switched off. Similarly, you would not need
a clock signal. There is still the RTC/CMOS RAM in the southbridge
"moat" of isolated circuitry, and it might handle things like
a front Power Button event as a PME and elevate the system to respond.

Some systems are designed to not respond to PME in S5 soft off,
others are still set up for that bit.

There is a "small economy and industry" running inside a PC when
you're not using it. I didn't even mention the Intel Management Engine,
where it is possible to wake a computer in an Enterprise setting
and push updates to it, or, the Intel ME can reboot a crashed
computer. With an ME and the correct bits and pieces, any PC
where the power switch is on at the back, and there is any sort of
network connection (the dual-ported Intel NIC), you can remotely
access the computer and do stuff to it. The kind of stuff "T"
dreams of :-)

When you buy a refurbished computer with an ME
inside it, one of the refurbishment steps at a place like
JoySystems, is to flash up the BIOS and remove any corporate
ME support so the machine can't be accessed that way. The ME
never stops running, and it's fed a "null code" to run instead
of a Minux OS. The ME processor is inside the Intel chipset
(PCH/Southbridge).

The status of an ME, is never re-assuring. My Optiplex 780 with
Q45 chip in it, the "Q" means "I have an ME", and if I run the
utility that checks the status of it, it returned "Disabled".
But it's never really disabled. To operate LoJack, it keeps
running for purposes like that. So while we hope the useless
utility that returns the status is being honest, from a hardware
perspective, we can never really be sure what is going on
inside the Q45. Each generation may get a new Southbridge/PCH
design, but there is generally one member of the chipset
family with the Q letter attached to it. And for networking,
there are Intel NIC for ME, and Intel Wireless for ME (so
an ME can be awakened over Wifi).

Paul

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 2025-05-18 20:24, Paul wrote:

On Sun, 5/18/2025 5:58 AM, Carlos E. R. wrote:

On 2025-05-18 11:32, Frank Slootweg wrote:

Carlos E. R. <robin_listas@es.invalid> wrote:

On 2025-05-17 19:51, Frank Slootweg wrote:

Carlos E. R. <robin_listas@es.invalid> wrote:

On 2025-05-16 16:35, NY wrote:

On 14/05/2025 12:09, Frank Slootweg wrote:

The PC doesn't wake on keyboard or mouse movement, and only resumes when
the power sort switch is pressed. When I put it to "sleep" (as Windows >>>>>>> calls it) there is period of about 10 seconds while it seems to dump the
RAM to HDD but when I resume it (by pressing power button) the logon >>>>>>> screen appears too quickly to have read from HDD.

Hybrid mode. If the battery runs out, it can recover from disk when >>>>>> powered.

    Hybrid Sleep. But I'm sure the OP ('NY') has a 'desktop', not a >>>>> laptop, i.e. no battery.

Still, they support sleep, using the PSU.

   Yes, as they should. Somewhere in this thread there's a suggestion >>> that a sleep of a 'desktop' shuts down all power. That's of course not
true, but as I don't have a desktop, I can't confirm.

It needs power, at least on the RAM and some chips to detect a
keyboard press and activate wakeup. I do not know if CPU is unpowered.

You would need Mechanical Off to remove all power.

Otherwise, if the PSU is switched on at the back, the +5VSB and
related circuits provide supervisory power.

To make a functional USB wake event, VBUS must be powered by +5VSB
so that the keyboard or mouse has +5VSB on it. The logic circuits
on board, some of those run off 3VSB or +5VSB. Things like PME
(Power Management Event) signal, those enter the logic as an indication service is required.

During Sleep, DRAM runs off +5VSB. The DRAM are in "AutoRefresh",
a kind of instruction useful for parking them. The DRAM have a
signal from the clockgen, so the chip is clocked. The voltage to
the DRAM can be slightly lower. (The voltage to the DRAM typically
comes from a small switching converter, powered by +5VSB). This
means the DRAM has to do a logical-OR of power sources. The DRAM can
also be a major drain (4W or 7W on a desktop, not a small amount
when the max output is 15W from the ATX supply on +5VSB).

If you have Wake On LAN enabled, the Ethernet chip may receive
core power so it can "listen". The receivers on the PHY likely
take a lot less power, than transmitting packets with 50 ohm drive
to the Ethernet isolation transformers in the black epoxy pack.
Switching off the Ethernet (because most people are not set up
for WOL Wake On LAN) is best but might only save you a watt.

I most check that next time I open the bios config.

In Hibernate, the converter for the DRAM does not have to be
enabled and can be switched off. Similarly, you would not need
a clock signal. There is still the RTC/CMOS RAM in the southbridge
"moat" of isolated circuitry, and it might handle things like
a front Power Button event as a PME and elevate the system to respond.

Some systems are designed to not respond to PME in S5 soft off,
others are still set up for that bit.

There is a "small economy and industry" running inside a PC when
you're not using it. I didn't even mention the Intel Management Engine,
where it is possible to wake a computer in an Enterprise setting
and push updates to it, or, the Intel ME can reboot a crashed
computer. With an ME and the correct bits and pieces, any PC
where the power switch is on at the back, and there is any sort of
network connection (the dual-ported Intel NIC), you can remotely
access the computer and do stuff to it. The kind of stuff "T"
dreams of :-)

When you buy a refurbished computer with an ME
inside it, one of the refurbishment steps at a place like
JoySystems, is to flash up the BIOS and remove any corporate
ME support so the machine can't be accessed that way. The ME
never stops running, and it's fed a "null code" to run instead
of a Minux OS. The ME processor is inside the Intel chipset (PCH/Southbridge).

The status of an ME, is never re-assuring. My Optiplex 780 with
Q45 chip in it, the "Q" means "I have an ME", and if I run the
utility that checks the status of it, it returned "Disabled".
But it's never really disabled. To operate LoJack, it keeps
running for purposes like that. So while we hope the useless
utility that returns the status is being honest, from a hardware
perspective, we can never really be sure what is going on
inside the Q45. Each generation may get a new Southbridge/PCH
design, but there is generally one member of the chipset
family with the Q letter attached to it. And for networking,
there are Intel NIC for ME, and Intel Wireless for ME (so
an ME can be awakened over Wifi).

My laptop is designed for business, but I think it doesn't have that.

--
Cheers,
Carlos E.R.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On Mon, 5/19/2025 6:38 AM, Carlos E. R. wrote:

On 2025-05-18 20:24, Paul wrote:

On Sun, 5/18/2025 5:58 AM, Carlos E. R. wrote:

On 2025-05-18 11:32, Frank Slootweg wrote:

Carlos E. R. <robin_listas@es.invalid> wrote:

On 2025-05-17 19:51, Frank Slootweg wrote:

Carlos E. R. <robin_listas@es.invalid> wrote:

On 2025-05-16 16:35, NY wrote:

On 14/05/2025 12:09, Frank Slootweg wrote:

The PC doesn't wake on keyboard or mouse movement, and only resumes when
the power sort switch is pressed. When I put it to "sleep" (as Windows >>>>>>>> calls it) there is period of about 10 seconds while it seems to dump the
RAM to HDD but when I resume it (by pressing power button) the logon >>>>>>>> screen appears too quickly to have read from HDD.

Hybrid mode. If the battery runs out, it can recover from disk when >>>>>>> powered.

     Hybrid Sleep. But I'm sure the OP ('NY') has a 'desktop', not a >>>>>> laptop, i.e. no battery.

Still, they support sleep, using the PSU.

    Yes, as they should. Somewhere in this thread there's a suggestion >>>> that a sleep of a 'desktop' shuts down all power. That's of course not >>>> true, but as I don't have a desktop, I can't confirm.

It needs power, at least on the RAM and some chips to detect a
keyboard press and activate wakeup. I do not know if CPU is unpowered.

You would need Mechanical Off to remove all power.

Otherwise, if the PSU is switched on at the back, the +5VSB and
related circuits provide supervisory power.

To make a functional USB wake event, VBUS must be powered by +5VSB
so that the keyboard or mouse has +5VSB on it. The logic circuits
on board, some of those run off 3VSB or +5VSB. Things like PME
(Power Management Event) signal, those enter the logic as an indication
service is required.

During Sleep, DRAM runs off +5VSB. The DRAM are in "AutoRefresh",
a kind of instruction useful for parking them. The DRAM have a
signal from the clockgen, so the chip is clocked. The voltage to
the DRAM can be slightly lower. (The voltage to the DRAM typically
comes from a small switching converter, powered by +5VSB). This
means the DRAM has to do a logical-OR of power sources. The DRAM can
also be a major drain (4W or 7W on a desktop, not a small amount
when the max output is 15W from the ATX supply on +5VSB).

If you have Wake On LAN enabled, the Ethernet chip may receive
core power so it can "listen". The receivers on the PHY likely
take a lot less power, than transmitting packets with 50 ohm drive
to the Ethernet isolation transformers in the black epoxy pack.
Switching off the Ethernet (because most people are not set up
for WOL Wake On LAN) is best but might only save you a watt.

I most check that next time I open the bios config.

In Hibernate, the converter for the DRAM does not have to be
enabled and can be switched off. Similarly, you would not need
a clock signal. There is still the RTC/CMOS RAM in the southbridge
"moat" of isolated circuitry, and it might handle things like
a front Power Button event as a PME and elevate the system to respond.

Some systems are designed to not respond to PME in S5 soft off,
others are still set up for that bit.

There is a "small economy and industry" running inside a PC when
you're not using it. I didn't even mention the Intel Management Engine,
where it is possible to wake a computer in an Enterprise setting
and push updates to it, or, the Intel ME can reboot a crashed
computer. With an ME and the correct bits and pieces, any PC
where the power switch is on at the back, and there is any sort of
network connection (the dual-ported Intel NIC), you can remotely
access the computer and do stuff to it. The kind of stuff "T"
dreams of :-)

When you buy a refurbished computer with an ME
inside it, one of the refurbishment steps at a place like
JoySystems, is to flash up the BIOS and remove any corporate
ME support so the machine can't be accessed that way. The ME
never stops running, and it's fed a "null code" to run instead
of a Minux OS. The ME processor is inside the Intel chipset
(PCH/Southbridge).

The status of an ME, is never re-assuring. My Optiplex 780 with
Q45 chip in it, the "Q" means "I have an ME", and if I run the
utility that checks the status of it, it returned "Disabled".
But it's never really disabled. To operate LoJack, it keeps
running for purposes like that. So while we hope the useless
utility that returns the status is being honest, from a hardware
perspective, we can never really be sure what is going on
inside the Q45. Each generation may get a new Southbridge/PCH
design, but there is generally one member of the chipset
family with the Q letter attached to it. And for networking,
there are Intel NIC for ME, and Intel Wireless for ME (so
an ME can be awakened over Wifi).

My laptop is designed for business, but I think it doesn't have that.

Check and see what your PCH is.

In the CPU-Z utility, under "Mainboard" tab, fifth row down, is
the entry for "Southbridge". You could check there for a
"Q" identifier.

I don't know if that identifier, is included in the output
from INXI. For some things like that, it uses DMI, and the
DMI might not have the detail needed.

Paul

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 2025-05-19 18:17, Paul wrote:

On Mon, 5/19/2025 6:38 AM, Carlos E. R. wrote:

On 2025-05-18 20:24, Paul wrote:

On Sun, 5/18/2025 5:58 AM, Carlos E. R. wrote:

On 2025-05-18 11:32, Frank Slootweg wrote:

Carlos E. R. <robin_listas@es.invalid> wrote:

...

The status of an ME, is never re-assuring. My Optiplex 780 with
Q45 chip in it, the "Q" means "I have an ME", and if I run the
utility that checks the status of it, it returned "Disabled".
But it's never really disabled. To operate LoJack, it keeps
running for purposes like that. So while we hope the useless
utility that returns the status is being honest, from a hardware
perspective, we can never really be sure what is going on
inside the Q45. Each generation may get a new Southbridge/PCH
design, but there is generally one member of the chipset
family with the Q letter attached to it. And for networking,
there are Intel NIC for ME, and Intel Wireless for ME (so
an ME can be awakened over Wifi).

My laptop is designed for business, but I think it doesn't have that.

Check and see what your PCH is.

In the CPU-Z utility, under "Mainboard" tab, fifth row down, is
the entry for "Southbridge". You could check there for a
"Q" identifier.

I only have Linux on it.

I don't know if that identifier, is included in the output
from INXI. For some things like that, it uses DMI, and the
DMI might not have the detail needed.

--
Cheers,
Carlos E.R.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)