Custom Windows 7 PC Won’t Shutdown

Topic

New Reply

Hello, Since 2018 or 2019 I’ve had this issue happen but not happen all the time or take less times to finally stay shutdown after trying tricks given to me here awhile ago. In 2018 or 2019 spent a lot of time with Microsoft Tech Support Level 2 personnel. They were able to fix worked for awhile.

But, when I started following instructions to bring my Win7 Pro up to date it required doing multiple restarts which triggered this shutdown issue. The physical shutdown button isn’t broken.

I wonder if upgrading to Win 10 will fix this issue?

Since Adobe stopped supporting Win 7, I’m forced to upgrade to Win10. Should this issue be fixed before upgrading???   I don’t want upgrading make it harder to finally fix my issue.

I have Custom PC Windows 7 Pro (x64) on SSD and have HDD.

Reply | Quote

Replies

I wonder if upgrading to Win 10 will fix this issue?

Probably it will. I recommend to do fullbackup befor attempting to upgrade. You can use Macrium Backup SW, for example, download the Macrium Reflect 7 FREE version at the “Backup at home” section.
What does take long to shut down? Is Windows closing long?
If Windows takes long to shut down, it seems like there is SW problem – some service refusing to shut down probably.

If the Windows closes and then it takes long to shutdown (black screen, power LED is on), then it couldd be HW issue.

Dell Latitude 3420, Intel Core i7 @ 2.8 GHz, 16GB RAM, W10 22H2 Enterprise

HAL3000, AMD Athlon 200GE @ 3,4 GHz, 8GB RAM, Fedora 29

PRUSA i3 MK3S+

Reply | Quote

Windows doesn’t take long to shutdown at all. Windows will boot back up again/my PC tower will turn on again either right away or 15 mins or 30mins or 1hour later. When I’m done using my PC and want shutdown/turn off the PC for the day or back when I was bringing it up to date and certain groups of KB patches and etc required restarting my PC a few times.

If upgrading doesn’t fix this issue, will it make it worse..harder to fix??

(My PC had Win10 then I quickly downgraded to Win7 pro) Will backing up save the Product Key of Win7 Pro?

Backup can be saved to my 4TB external HDD that I use and if I end up needing to use it, I just plug it in and open the backup file?

Using the backup feature that comes with Win7 is an option too, right..

Reply | Quote

TheFamilyIT wrote:

Windows doesn’t take long to shutdown at all. Windows will boot back up again/my PC tower will turn on again either right away or 15 mins or 30mins or 1hour later.

So it does shut down, but then wakes up sometime later while unattended?

If that’s the case, it could just be caused by vibrations to a sensitive mouse.

Reply | Quote

Really..the mouse could be issue? My PC desktop has been sitting in the same place and using same mouse since purchased since 2014. And this problem didn’t start any time close to 2014. I never manually set up my mouse to be able to turn on my PC.

How do I stop the vibrations?

Reply | Quote

Does the problem occur immediately after a boot?

Do you have any network devices that are not available?

Have you checked Task Manager to see if any process is taking over?

What other steps have you taken to diagnose the issue?

--Joe

Reply | Quote

The problem does not occur right after I turn on my PC. My router works..should be available since I have no problem using internet. What other network device are you thinking of?

How do I tell if there’s a process taking over??  And what do you mean by taking over?

My PC will boot back up again/my PC tower will turn on again either right away or 15 mins or 30mins or 1hour later. When I’m done using my PC and want shutdown/turn off the PC for the day or back when I was bringing it up to date and certain groups of KB patches and etc required restarting my PC a few times.

I can’t remember all what Microsoft Tech support did way back then. I just remember something to do with hibernation issue. But, the other steps/tricks that I take to turning off PC by pressing the physical on/off button of PC tower or alt+ fn4 or going into power settings and restoring the balanced settings to default or powerconfig.exe hibernation off command, clicking shutdown from start menu and quickly hold down physical on/off button. Only I f necessary step I was told, is use an elevated command prompt shutdown /s/f/t 00

Finally step/trick is to physically unplug my PC’s power from wall (which hard to do back when was window updating my PC & problem was bad since it’ll turn on again faster than I can I could have unplug)

When I done using the PC, I try a different step/trick listed above (starting with the least invasive/simple one) each time my PC turns itself back on until hopefully one works. The first step/trick I try is using physical button and then I keep checking up on my PC to see if it’s stayed turned off.

 

Reply | Quote

I agree that Windows 10 has a chance of fixing this, and if you want to upgrade anyway I would do that first.  If the problem continues, the first thing I would try is to use wireless internet instead of wired.  After that I would suspect the mouse as above but I am not sure how to change settings to prevent mouse wake up.

Are you sure you need the support from Adobe?  Do you want to discuss which Adobe software it is?

Reply | Quote

You write “right away or 15 mins or 30mins or 1hour later”. Does this mean that the computer turns itself on again: immediately; or at exactly one of these intervals, or do you mean that the turn-on time is variable?
If it is variable, about what is the shortest interval (the quickest), other than ‘immediately’? And maybe what is the longest interval?

Some things to check –
1. Right-click on: Start orb > Properties > (tab) Start Menu.
What is the setting for the ‘Power button action:’?
Note: This ‘Power button’ is the text, and action, that is at the bottom of the right section of the Start Menu.

2. Note: The ‘Power button’ below is the physical power button on the computer itself.
What is the setting at:
Control Panel > Power Options > “Choose what the power button does” > “When I press the power button”
Note: Even if this is set to “Shut down” it may not completely solve the problem. But let’s begin here.

3. Control Panel > Device Manager > Network adapters > (rt. click on the adapter, which gives) > Properties > (button) Change settings > (tab) Power Management.
This is on my Win 7 tower:

Repeat this verification if you have more than one network adapter, such as wireless.

The Power Management setting for the Mouse (use same pathway as above) probably DOES ‘Allow this device to wake the computer’. This is normal.

Are there any other ‘input’ devices on the computer, such as a Game attachment?

Reply | Quote

It is at variable times when my PC turns back on. How many times it takes for one of the tricks/suggestions that I’ve been told in the past to do, it varies every time I go to use my PC.

For example, today, 1st try I did the suggestion(restoring to default power options balanced settings ->Alt+fn4 ->shutdown) that worked last time. But, turned back on in 1 min then on the 2nd try using another trick(using physical on/off button) my PC stayed off.

I checked everything that was mentioned:

1) The setting for the Power button action is Shutdown

2) It is set to Shutdown

3) My Tab doesn’t have the same options as in your screenshot. Properties window had more tabs too. I couldn’t find “Allow computer to turn off this device to save power”

For the Mouse, 3 was listed and 1st one listed says HID – Compliant Mouse and after right-clicking to check its Power Management settings. I see both settings  “Allow this device to wake the computer” & “Allow to turn off this device to save power” are disabled.

2nd one listed says HID – Compliant Mouse and it’s Pwr Management Settings is the same as the 3rd one listed (mentioned below)

3rd one listed says Microsoft P5/2 Mouse and it’s Pwr Management Settings is “Allow this device to wake the computer” & “Allow to turn off this device to save power” is grey out/grey text

When looking at Device System Manager, the only other “input” devices on the computer is Wacom Cintiq drawing display. Does that count…

Reply | Quote

I forgot in my reply to attach screenshot of how my Network Adapter Properties Window’s Power Management Tab look like

Reply | Quote

Uncheck (disable) the Wake On LAN settings shown in your screen capture. You also might want to disable the Wake on ARP and NS requests. After all, a computer which is shut down shouldn’t respond to anything.

Reply | Quote

Hello TheFamilyIT,

You mention that it is a custom Win7 computer. What is the brand and model number of your custom computer’s motherboard? I ask, since I think that we might need to look at settings in your motherboard’s BIOS. Perhaps in 2018 or 2019 you replaced the battery for the BIOS and the BIOS reverted back to its default settings?

How old is this computer?

Is this computer merely on your home computer network? If yes, I assume that you are not remotely administering any of the computers by remotely forcing them to wake in order to install updates or programs? If yes to this last question, then please look at the installed programs. Is Intel Management Engine Components installed? If yes, what version and when was it installed?

If you ARE NOT remotely administering any of the computers by remotely forcing them to wake in order to install updates or programs, and you are not doing any kind of remote Internet access which requires waking up the computer, then uninstalling the Intel Management Engine Components may the first order of business since years ago a very serious security vulnerability was discovered in the Intel ME drivers! The point is that a hacker could be waking up your computer, or that another device on your home network could be waking up your computer.

Another possibility is that wake up features are turned on in your computer’s BIOS. That is why, at the very top, we might need to look at your computer’s BIOS settings. We need to know what the make and model is of your computer’s motherboard in order to see the default settings for your computer’s BIOS, and to point you to what settings to check and possibly change in the BIOS.

Okay, so now you have an idea about the info we need, and the possible things which we might ask you to check. People have offered suggestions. Instead I would like to get to the root of the problem.

Best regards,

–GTP

Reply | Quote

Motherboard brand and model: Gigabyte Ultra Durable Motherboard GA-X99-UD4 (manual edition date stamp August 2014)

The only battery replaced might have been that big round battery for the clock/time seen on screen after having PC unplugged for 5 days. (Happened long time ago/can’t remember when).

My  computer was built in 2014. Yes, it is only on my home network. The only remote thing is when Microsoft Tech support or some other support is helping me. I have recently tempted to do is remotely access my software by using a non-powerful laptop so I can work on stuff anywhere in the house. Would this mean that I should keep Intel management Engine Components?

Reply | Quote

Check the settings in your motherboard’s BIOS according to my screen captures of your GA-X99-UD4 motherboard manual.

Reply | Quote

I don’t want to accidentally mess anything up in the BIOS, could you give me detailed step by step on how to get to each of those settings (including how to get to the menu version/style that has chipset and power management settings), please?

I press quickly the delete key on keyboard before Motherboard’s logo screen disappears, right??  And is that the only way to open the BIOS (just in case I’m not fast enough) ?

Also, if in the future I want to remotely access my softwares by using a non-powerful laptop so I can work on stuff anywhere in the house. Would any of these settings would need to be changed?

Reply | Quote

Hi TheFamilyIT,

Here is the link for the PDF of the product manual for your Gigabyte GA-X99-UD4 motherboard:

https://www.gigabyte.com/Motherboard/GA-X99-UD4-rev-11/support#support-manual

Earlier I posted screen captures from the PDF. For reference, I have attached those screen captures again.

It turns out that your motherboard has two BIOS chips. One serves as a backup BIOS in case you actually manage to seriously mess up the settings in the primary BIOS. That is pretty slick. The more I look at the specifications and features of your motherboard, the more I realize just how really nice this motherboard was for the year 2014.

To get into the BIOS, simply start tapping the Del key immediately after you power on your computer. Chapter 2 of the PDF is all about the BIOS settings. Chapter 2 starts on page 37 in the PDF. Gigabyte does a good job of explaining each of the BIOS settings. If you want to learn more about any particular BIOS setting, simply google it.

Don’t worry about messing up any BIOS settings since you actually have to choose to save any changed settings before you exit the BIOS and restart the computer. If you somehow manage to mess up the BIOS settings such that your computer won’t boot, simply follow the instructions on page 35 of the manual which are listed under 22) CLR_CMOS (Clear CMOS Jumper).

It is a good idea to familiarize yourself with what types of settings are grouped together under each of the main menus. Gigabyte does a good job of showing what settings are under each menu and listing the default value for each setting.

Best regards,

–GTP

 

Reply | Quote

Manual doesn’t say how to get to PCH SATA Configuration settings or the SATA Mode Options where those specific settings (in a beginner/user friendly way) that you want me to check.    Manual’s screenshot doesn’t show those settings.

Manual doesn’t say what to click inside Startup Guide to get to the other Modes/other guide formats, like ST Mode and Classic Setup(F2 is the only thing that says Classic Mode), BIOS Setup Menus(manual says settings that I need to check are in bios setup menu).

Reply | Quote

These settings are under the Chipset settings menu. See page 60 in the manual for how to configure these settings.

The most important option is whether or not all detected SATA devices are to be configured as operating in either IDE or ACHI mode. I am nearly 100% sure that this setting should be ACHI mode.

Note that the other settings normally are automatically configured by the BIOS when the computer boots up. The only changeable settings are for whether or not the sSATA device port is enabled (all should be enabled), whether or not Hot Plug should or not be enabled if the hard drive plugged into the given sSATA port is a given device actually is a removable hard drive, and what type of device the sSATA device is (hard drive or whatever the other options are).

As far as your second paragraph, I need to dig into the manual’s modes and settings for your motherboard BIOS.

In any event, try to find and change the specific settings for which I previously posted screen captures. After making sure that those were the only settings which you changed, choose to save those settings in BIOS and reboot. If those are the only settings which you actually change and save, then there is absolutely no way that those changed BIOS settings will brick your computer, and that there is a good likelihood that these changed settings will resolve your issues.

Best regards,

–GTP

Reply | Quote

Hello GTP, You are a good man. You have some very good ideas and are very thorough.

If you and Ascaris do not make headway with FamilyIT guys auto starting computer, see my post #2385550.

Reply | Quote

Hello @GoneToPlaid , have you seen my last post(from the 23rd)asking for navigation (in detail) to BIOS settings, by any chance?

(just trying out to see if tagging is a an option in forums)

Reply | Quote

There seems to be some confusion about whether the problem is the PC waking up (coming out of standby) or acting like it is turning itself on. The advice about unchecking the box for “allow this device to wake the computer” only applies to sleep, and only once it has booted into Windows.

The way I read it is that the computer is being shut down, not put to sleep, as in turning it off, but it doesn’t remain that way, but appears to turn itself on and boot into Windows uncommanded.

If that is correct, the OS is not even loaded at the point that it is coming back to life on its own. The PC is fully under the control of the UEFI or BIOS at that point. I’d say either there is some UEFI/BIOS setting to automatically turn on for some reason, or that there is a power issue (power to the unit is intermittently going out, or the power supply is having issues).

In the UEFI/BIOS, there is often an option for the PC to come on after a power loss. If so, and if you find it enabled, you might try turning that off, and if it fixes the issue, you know that you’re having power issues. The power supply could be failing, and even if the issue appears to be fixed, it’s better to get to the bottom of it now than to have something worse happen. Power supplies usually just stop working when they die, but sometimes they can take out motherboards and other devices in the system as they go.

If the power to the PC is being switched on and off for some reason, switching that option to not come on when power is restored should fix it, if for some reason you want the power to be turned off and on as it is.

If there is no option for that, it could be that it is the normal behavior of that motherboard. You can test this by unplugging the power with the PC shut down, then plugging it back in, and seeing what it does.

Dell XPS 13/9310, i5-1135G7/16GB, KDE Neon 6.2
XPG Xenia 15, i7-9750H/32GB & GTX1660ti, Kubuntu 24.04
Acer Swift Go 14, i5-1335U/16GB, Kubuntu 24.04 (and Win 11)

1 user thanked author for this post.

GoneToPlaid

Reply | Quote

No disagreement to this.
Are we absolutely positive that there is a true power-off shutdown?
Can sleep/hibernate ‘fake it’? The time delays are puzzling.
There has been no mention of a UPS: could/can a UPS trigger a power on?

Reply | Quote

Hi everyone,

Ascaris, very good observations! I too was thinking along these same lines of thought. Paul T also mentions checking the OS wake timers by using powercfg in an admin level command prompt. Anyway, I figure that it would be really helpful if TheFamilyIT guy would post info about his computer’s motherboard so that we could help him look at some of his motherboard’s BIOS settings which might be causing his computer to wake up from time to time.

PaulK asks if Windows 7 sleep/hibernate functions can possibly “fake it”. The answer is No. Yet what can occur, due to hard drive I/O overhead issues related to Sleep and Hibernate and which I won’t discuss at this time, is that Windows 7 can go into an improper shutdown state. Yet this would never cause Windows 7 to suddenly wake up since it would be similar to suddenly killing all power to the computer.

I figure that the issue might be in the motherboard’s BIOS settings. Yet I also have a concern that TheFamilyIT guy may also have outdated Intel ME drivers installed which have serious security issues. Perhaps you all will recall the “Intel ME bug” from years ago?

We also don’t know if TheFamilyIT guy’s Internet access is through a router with NAT, or through a simple modem with no NAT and no hardware firewall.

So aside from us helping TheFamilyIT guy resolve his computer’s waking up issue, I also think that we should have him look at his computer’s Windows event logs to see what is waking up his computer.

We should also consider that TheFamilyIT guy has not disabled Remote Assistance and Remote Desktop access to his computer, and whether or not he has blocked RDP port 3389, since we have no idea about the status of his installed Windows security updates.

We also don’t know if TheFamilyIT guy has installed remote access software such as RealVNC or GoToMyPC, or similar remote computer access software. Previous and perhaps current versions of such software are readily and remotely hacked.

There is so much about TheFamilyIT guy’s computer which we don’t know, in terms of its installed security updates, in terms of installed remote access software, and in terms of the motherboard’s BIOS settings.

I have always been a bit paranoid of any computer which magically wakes up for no apparent reason after performing a full shutdown, since there potentially are many security related things to check for and then to rule out.

Anyway, I am simply throwing my ideas onto the wall in order to see what might stick to the wall. This is all nothing more than “food for thought” in terms of things to think about while we try to help TheFamilyIT guy resolve his computer issues.

Best regards,

–GTP

Reply | Quote

Hello Ascaris, Nice to see you. I’m the guy you talked to years ago about plastic on engines (I am opposed). We had a great conversation!

I want to agree with you that this may be a BIOS event or other hidden “wake up function”. Something is waking it up from a USB (mouse) input, a timed event, a BIOS “wake on” event, etc.

FamilyIT guy should read your suggestions and investigate them. I’m sure you and GoneToPlaid will get some results.

His last choice may be my dirt cheap, fast idea, to put a strip surge protector on the desk beside the PC, plug the PC CPU “box” into that surge protector. When he shuts down the PC, flip OFF the surge protector. My Brother use to do that all the time to protect the PC from lightening strikes, and that works.

Hope this helps.

Reply | Quote

View System/Kernel-power related events in Event Viewer to see what is bringing the computer out of shut down.

Also check what events are configured to wake the computer using an admin Command Prompt: powercfg /waketimers

cheers, Paul

Reply | Quote

Where is Event viewer located? And once I view System/Kernel-power related events what do I look for/tell in what I see that means it’s bringing the computer out of shut down. (I don’t know how to your advice)

Reply | Quote

Press the Win key and type “event”. Without the quotes.

Look at the system events.

Let us know what you find.

cheers, Paul

 

Reply | Quote

There is a very long list with Information listed under Level column, bottom of list date stamp is from 2018 to today’s date. I don’t know what I’m looking for that will tell me something is wrong.

Under Source column, there a number of times the Kernel-Processor-Power, for today’s date and time(all with same time stamp and listed 14 times) I see event ID 26 task category (4) But then 2 more with Event ID 12 & 13 Task Category is None.    I can go on with more..

I also see an item listed under Level column “Error” with August 16 2021 (with the time of one of my attempts to keep my PC turned off.).  Source Column Schannel Event 36887  Under General Details says “the following fatal alert was received: 20

Under Level column, “Warning” with same date but with time 10 mins earlier than the “Error” Source e1 dexpress Event 27 General description says “Intel(R) Ethernet Connection (2) I228 – V Network link is disconnected”   There are 3-4 of this same “Warning” listed with same date and with times.

But, there is e1 dexpress with same date and 1 min earlier that doesn’t have “Warning” under Level Column General Description says “Network link has been established at 1Gbps full duplex”.  This happens/listed a couple times

After running the command power cfg/wake timers,  it says There are no active wake timers in the system.

Reply | Quote

When you select the power events you will see a reason in the lower pane. e.g.

The system has returned from a low power state.

Sleep Time: ‎2021‎-‎08‎-‎21T07:15:25.243238900Z
Wake Time: ‎2021‎-‎08‎-‎22T06:05:08.096450600Z

Wake Source: Unknown

If you have no wake timers in the system this usually means the system is being woken by a hardware event, e.g. pushing the power button or moving the mouse.

wavy wrote:

Have you checked your Bios for a ‘wake-on-lan’ setting?

This is an unlikely source of power up. It requires another device to send the PC a “wake up” packet on the network and these are always manually configured.

cheers, Paul

Reply | Quote

Paul T wrote:

This is an unlikely source of power up. It requires another device to send the PC a “wake up” packet on the network and these are always manually configured.

seems we are in unlikely territory already, no?
I am also wondering about the Intel Management Engine possibilities. Disable the BIOS level IP stack perhaps??

🍻

Just because you don't know where you are going doesn't mean any road will get you there.

Reply | Quote

Unlikely that it’s a WOL issue because you would know if you had set this up yourself via another device.

cheers, Paul

Reply | Quote

After running the command power cfg/wake timers,  it says There are no active wake timers in the system

Reply | Quote

Have you checked your Bios for a ‘wake-on-lan’ setting?

🍻

Just because you don't know where you are going doesn't mean any road will get you there.

Reply | Quote

I see this as neither a software issue nor a BIOS issue, but as an age issue.

I suggest a deep cleaning.  Shutdown the PC (by whatever means), unplug the power cord from the back of the PC and remove all other connecting device plugs.  Open the side of the case, and use some canned air to blow out any accumulated dust.

Then carefully disconnect/reconnect every motherboard connection at the motherboard, including the 24-pin power connector, fan connectors, etc.  After that step, do the same for the other ends of the cables (not the power cables from the PSU).

In other words, disconnect the power connector from your hard drive(s) and reconnect it and disconnect the data cable from your hard drive(s) and reconnect it.  If you have an optical drive, to the same to it.

After this procedure, put everything back together and power on your PC.  After it boots into Windows fully, do a normal shutdown and see if it stays shut down.

Always create a fresh drive image before making system changes/Windows updates; you may need to start over!

We all have our own reasons for doing the things that we do with our systems; we don't need anyone's approval, and we don't all have to do the same things.

We were all once "Average Users".

Computer Specs

Reply | Quote

Hello All,

I apologize for interfering into this thread, but I do encounter a very similar issue with my Windows 7 PC.
My PC is working without any problem at all for about 5 years, but, about 3 months ago, it suddenly started to “not being able to stay in sleep mode”.
I browsed the web for several days, tried out ALL the suggestions and ruled out ALL the possible software settings discussed in all the places and nothing helped.
After about 3 weeks of unsuccessful endeavors, one day I kept the machine busy all day long by performing various backups, large downloads from the Internet, a.s.o. and, the next day,
the entire problem suddenly went away and everything stayed ok for about 2.5 months !
It looks as if the machine was “tired enough” for not refusing to sleep 🙂
Then, one week ago, the same problem suddenly started again.
After another day of intensive work, the problem seemed to have solved itself again, but this time it only lasted for one day, then the problem started repeating again.

What happens is that, when the machine enters sleep mode, either due to 30 minutes of being idle or by manually choosing the “Sleep” action, then it suddenly starts again.
Sometimes this might happen after 1 or 2 hours, but mostly it happens after just a few seconds after entering sleep, as if the sleep mode was not even acquired completely.
There is no problem at all if I shut down the machine, it never started by itself after being shut down, it only starts by itself from sleep mode.

My feeling is that the problem resides somewhere in the power supplier, but I am not yet able to convince my computer technician about this.
Also, because the problem is repeating itself at random times and with random frequency, it might be hard to exemplify it at any given moment, so that my technician be able to see it happening.
Is there anything that can be done to prove that  the issue is indeed hardware-related
( specifically, power supplier related ) ?
Unfortunately, I am not familiar with hardware/mechanical issues, my knowledge
(and really do have a lot of it !) is only software related.

Thanks a lot in advance & Best Regards,
iudith.m

Reply | Quote

Have you tried the suggestions in #2384598 above?

cheers, Paul

Reply | Quote

Hello Paul,

Yes, of course I tried all these suggestions.
The Wake-up reason is NEVER logged correctly in the Event Viewer,
it always shows “Wake Source: Unknown”.
Interestingly, the Sleep reason is logged correctly, showing
“Sleep Reason: System Idle” if it enters sleep by itself or
“Sleep Reason: Application API” if I manually choose the Sleep option.

Executing “powercfg /waketimers” shows that
“There are no active wake timers in the system”.

Cheers and many thanks,
iudith.m

Reply | Quote

Try using a wireless mouse if you are using a wired one.  Wired mice can wake with very tiny shakes.

Reply | Quote

Hello,
The mouse was the first possible culprit we thought of, but the same happens even if I unplug the mouse.
Otherwise, any mouse movements are excluded when there is no one in the room for many hours …
Also, both the mouse and the Network adapter have
“Allow this device to wake the computer” set to “Disabled”.
(aka the only device allowed to wake up the computer is the Keyboard.)

When it wakes up immediately after entering sleep mode, my feeling is that the power does not go off completely, and then “something” is notifying the OS to resume working.
But this does not explain the situations when the wake up happens after several hours of sleep, but still without anyone touching anything.

Cheers and many thanks,
iudith.m

Reply | Quote

I have seen this back in my Win 7 days. I suggest that you review the following article.

https://www.ghacks.net/2013/12/31/find-pc-wakes-stop/

HTH.

Regards, Phil

Reply | Quote

Could you choose hibernate instead of sleep?  Hibernate might stay shut down.  The automatic and manual options can both be set as hibernate instead.  You will likely have to enable hibernate first.

https://www.intowindows.com/how-to-enable-hibernate-option-in-windows-7/

https://www.guidingtech.com/13592/change-windows-7-power-button-action-sleep-hibernate/

In my experience problems like yours have been hard to debug, requiring some luck and drastic options like o/s reinstall or upgrade.  Would your workload work well on Windows 10?  Free upgrade is typically still possible.

Reply | Quote

I suspect that one or more wake timers is enabled in your computer’s BIOS. Also, try disabling Intel Rapid Start Technology in BIOS.

Reply | Quote

Hello All,

This thread seems to still be very interesting, so, I would like to ask for your opinion and/or experience on what I am experiencing already for a few days. But, now the issue seems to have become kind of “deterministic”, which, in my opinion, should lead to a logical explanation.

As a reminder, my problem is as follows:

I have a custom Windows 7 PC, which was constructed by my technician back in Nov 2016, and it worked OK all the time.
At the beginning of Dec 2021, it started to experience random wake ups from Sleep, and I was not able to locate the source of the problem.
The “Wake reason” for ANY of wake-ups from Sleep is shown in the Event viewer as “Unknown”.
Using powercfg /lastwake and all the related powercfg options also do not show any useful information, regardless of whether I woke up the PC by pressing a keyboard key or if it was one of those random wake-ups.

My technician insisted there should be some software reason for the random wake-ups because Windows 7 is already old and so on.
Being a software person myself, I could not accept that software could cause any such behavior, except maybe if there was some strange malware or something, which was not the case.
Since I insisted on continuing to use Win7 (life is so, so easy without the MS monthly updates… and prayers), my technician could not give me any constructive help. My feeling was always that it was surely some power-related issue and purely hardware.
My idea was that maybe the PSU should be replaced, or at least checked, since many times a wake-up happened just a few seconds after entering sleep mode.

After about two weeks, the problem disappeared by itself, then appeared again about 3 months later for a few days, then disappeared again and everything was OK from Mar 2022 up to July 2023 !!!

At the beginning of July our internet infrastructure was changed, so my router was replaced and I started to encounter some strange warnings logged by the BITS-client service, followed by a buggy Windows Defender signature update from MS; which we discussed here in much detail in another thread. Luckily that problem corrected itself.

Less than a week later, I again started to encounter random wake-ups and, kind of logically, I could only think it was some networking related issue following the router change. But, after again going through all the software-related knowledge available, I found no one to help me; or even listen to me.

So, as a last attempt, I unplugged both the keyboard and the mouse (the mouse being an almost new one) which were plugged into the PC case’s rear USB slots just to see what happened. And here’s the miracle. Once I did that, all the random wake-ups stopped happening. I worked on another computer connected to the same home network and, for about 2 days, there was not a single random wake-up.

Then to move forward, I decided to try plugging my keyboard and mouse into two of the USB slots on the front of the PC thinking maybe that would work? There are 4 USB ports on the front, but only one of them worked with the keyboard while the mouse worked with 3 out of the 4 (the 4th one is USB 3.0 but, as far as I understand, that should not prevent a device from working, except maybe for speed issues a.s.o.)

This, however, raised another question in my mind:
Why are these USB ports not working as expected as they’re pretty much unused except for occasional backups to an external hard disk?!?

And now the “cherry on the cake top” :
I thought things are somehow up and running, and indeed, there were no more random wake-ups. But, can you guess what?
Switching the light in the room on or off was enough to trigger a computer wake-up!!!
So this time, it’s NOT a random wake-up, but one triggered by a completely unrelated cause!
And yes, it only happens when the keyboard & mouse are connected; if I unplug them, there’s no wake-up triggered by the light switch!

I browsed the web and found others have also asked this question, sometimes many years ago, so I’m not that stupid… because it does indeed sound crazy!

This definitely strengthens my original hypothesis that, somehow, the PSU is influenced by “some external cause” and it acts crazily, though now the issue seems very closely related to “wrongly powering the USB’s”.

From there, I came up with another idea:
My computer has two USB Root hubs. One shown as USB 2.0, which I guess is for the rear USB ports, and the other one as USB 3.0 which is probably for the front ports because I already know I have a USB 3.0 port on the front.
So my idea was, maybe I could try unchecking the “Allow the computer to power off this device” for the USB 3.0 root hub where my keyboard and mouse are connected and see what happens.
I thought, maybe if the USB 3.0 hub is not powered off, it’ll be “less sensitive” to power fluctuations; which is the only thing I can somehow relate to switching the light on/off?!?

Guess what?
This had no influence at all, the switching light on/off continued to wake up the PC 🙁
But, if I changed the same Device Manager setting for the USB 2.0 root hub (the hub where no device is currently connected) it seems to stop the wake-up issue :):)

To me, it’s pretty clear the USB’s are not working correctly; probably because of some PSU problems!
Now it all looks clearly deterministic and, if you ask me, my first attempt should be to replace the PSU and, if possible, try to replace the USB’s .

I’m not a hardware specialist, but I think I did correctly analyze the problem for my technician up to a “ready to go” state where he should just use his tools without touching any software or the OS in any way. The simple fact that my OS is an older version is NOT a reason for it not to work.

The problem is that MS makes it so, so hard to replace any piece of hardware, or maybe the entire machine, if the hardware is worn out 🙁 🙁
Of course this unfair behavior of MS is a topic for another thread… and there are many like that on this and other forums.

A last technical question I’d like to ask is whether it’s OK to use USB ports on an external USB hub for plugging in the keyboard and the mouse?
All this of course with the idea at least one of the existing USB ports can be made to work OK.

By the way, today I used the USB 3.0 port on the front for an external hard disk and it worked OK, although this same port did not “accept” either the mouse or the keyboard.

Now I can only hope I can convince my technician to give up the theory that Windows 7 is the culprit and finally accept my idea that the PSU and/or USB ports are the real problem. Once again, I’m sorry my hardware knowledge is far from my software one so I still need physical help to solve such issues.

I apologize for this long post and thank you for the patience to go through it and maybe share some ideas or tips that could be helpful… and convincing 🙂

Thanks a lot & Best Regards,
Iudith Mentzel

Reply | Quote

Not even considering the cost of technician time, either a power supply or a motherboard (the parts that likely contribute to USB issues) are expensive parts.  The preferred solution for this would usually be a workaround, like finding which USB port can be used with a hub to avoid the problem.

There is another part that could be involved – the power cords and surge protectors for the system.  As a test, remove all UPS and fancy surge protectors and use a simple surge protector with the minimal number of devices plugged in such as only computer and monitor, also replacing both those power cords.  This should prevent “ground loops” and may fix the problem.

Reply | Quote

Unfortunately, my optimism was too early 🙁
After apparently fixing the problem by changing the USB root hub settings, it however came back …
Then, as you say, I tried various combinations of sockets for the PC power … but, again, it only apparently worked …
I tried plugging the PC power cord both with and without a surge protector, but ultimately both behave the same.
I usually use surge protectors only for my monitors, not for the PC itself.

For me it is clear anyway that the power that the PSU receives when switching the light on/off causes it to wake up the PC, probably by “activating” the USB.

I also tried to unplug/replug the power cord, and now I am trying another power cord.
Unfortunately, for dealing with the real culprit I need the technician …

Changing the PSU is not that expensive … the motherboard is more problematic, because
the OS is “tied” to it …

I have several devices plugged in the same room, in adjacent power outlets: computers, monitors, router, but only this specific PC is sensitive to the light switch issue.

I have seen that there exist sets of power outlets with 8-10 outlets, having the surge protection already incorporated into them.
I wonder whether such a device could be helpful.

Thanks a  lot 🙂

Reply | Quote

Here’s a wild guess.
Check out @Myst ‘s information in this thread.

Reply | Quote

That’s interesting.
But, in my case my mouse is a wired one, at it is defined as NOT allowed to wake up the computer.  And, indeed, it does not !

The only device allowed to wake up the computer is the keyboard,
and the powercfg commands correctly show this.

As long as the wake-ups were completely random, I also suspected that the mouse could be a culprit, and then some task running … which in fact was not !

Now, the very precise correlation between the wake-up and pressing the light switch on/off
demonstrates that it is a power issue involved, which we should find.

What drives me crazy is that the OS never logs correctly the reason for the wake-ups,
it always says “Unknown” …

Reply | Quote

iudith.m wrote:

Switching the light in the room on or off was enough to trigger a computer wake-up!!! So this time, it’s NOT a random wake-up, but one triggered by a completely unrelated cause! And yes, it only happens when the keyboard & mouse are connected; if I unplug them, there’s no wake-up triggered by the light switch!

Any chance the light bulb controlled by that switch is a “smart bulb” (i.e. it can be controlled from a phone or via WiFi?)

If so, when it’s first powered, on it’ll send out a fairly strong RF signal in an “attempt” to establish a remote connection and that could explain why switching it on/off causes the PC to wake-up with the keyboard & mouse are plugged in (i.e. the RF signal is strong enough and close enough that it “bleeds over” into those devices making the PC think they’re actively being used.)

Reply | Quote

No, the light in the room is a completely regular old one, with a simple wall on/off switch.
Nothing smart or fancy.

Reply | Quote

@iudith.m, Very interesting indeed. Good testing on your part. My tech support person mentioned that a possibility to the wake up with lights being turned on could be due to small power fluctuations. A “good” surge protector with line filtering (to get rid of erroneous signals) should guard against this happening. If you have a powered hub, best to put it in the surge protector as well. If you exhausted all other possibilities, then maybe the power supply is acting up.

Reply | Quote

I would like to ask whether it is recommended to use a set of several power outlets, with surge protection incorporated in the set itself?

I have several devices (computers, monitors, router, and maybe later, a powered USB hub ), but up to now I’ve used individual power surge protectors only for my 2 monitors.

Thanks a lot 🙂

Reply | Quote

Hello Judith, Yes, You definitely also need surge protection on the PCs. I see no advantage in using different outlets. Suggestions on high quality surge protection are Tripp Lite ISOBar6Ultra or Brooks Surge and Noise VCN6-6 (I currently use these). If you decide to pursue this route, perhaps your technician can provide more advice. High quality surge protectors typically will indicate if there is a grounding fault.

Reply | Quote

I definitely intend to have a power surge protector.
After looking into this issue, I discovered power surge protectors also have a kind of “planned limited life” so now, on the two power surge protectors that I was using for my monitors, the green light that says “protected” is currently off and only the red light is on, showing there’s power.
So, when I tried to use one of those just as a test to connect my PC and see whether a surge protector could correct the problem, the test was meaningless.
If I remember correctly, in the past the green lights used to be on so this indeed might be a sign of “aging” and I should replace them. But, if I’m wrong, then maybe I never had “real” power surge protection and that could be an even bigger issue.

The problem with all these electrical devices is that we need to buy them locally otherwise they might not fit our local outlets… which are by themselves extremely variable from one case to another. So this reduces the “choices” to only devices that are available locally.
I’m now after a “box” (set of 8 outlets), which has surge protection built into it as a whole so I’ll be able to plug in all my devices into it.

What I don’t understand though is why turning the room light on/off only affects one of my computers and not the other two?
I’d guess some kind of power surge protection (or at least the voltage protection part of it) should be built into the PSU of the computer itself, regardless of whether I use an external outlet protection device or not.

Reply | Quote

I disabled all wake triggers in my computer’s BIOS. I also don’t allow any scheduled tasks to wake my computer. The only way that I can wake my computer from sleep or hibernation is to momentarily press the computer’s power button.

Perhaps try disabling the remote desktop and remote assistance services, and then block all inbound traffic on port 3389 in Windows Firewall or in whatever firewall program you are using on your computer. The reason is that it could be a hacker trying to get into your computer. Also check your router’s firewall and make sure that network address translation NAT) is enabled.

Additionally, use an online scanner for malware or PUPs on your computer. It has been years since I have had to deal with any malware, so please ask others about what online malware scanners to use on your computer.

Reply | Quote

Perhaps try disabling the remote desktop and remote assistance services, and then block all inbound traffic on port 3389 in Windows Firewall or in whatever firewall program you are using on your computer. The reason is that it could be a hacker trying to get into your computer. Also check your router’s firewall and make sure that network address translation NAT) is enabled.

Additionally, use an online scanner for malware or PUPs on your computer. It has been years since I have had to deal with any malware, so please ask others about what online malware scanners to use on your computer.

Reply | Quote

Hello All,

Just to summarize the current conclusions after several days of testing:

If the mouse and keyboard are plugged into the back side USB’s, I experience random wake-ups from sleep while only the keyboard is allowed to wake up the computer. As a special mention, these wake-ups are NOT related in any way to turning the light on/off in the room!

If the mouse and keyboard are plugged into the front side USB’s, the random wake-ups DO NOT happen any more!
Instead, the computer wakes up DETERMINISTICALLY each time the light is turned on/off in the room.
If I disable the keyboard (and also the mouse) from being allowed to wake-up the computer, there are no more wake-ups turning light on/off in the room! In this case, I can only wake up the computer by pressing the power button on the front of the case… not the nicest thing to do.

So, to me, it’s clear the two different USB hubs (on the back side and the front side) are “treated” differently by the PSU and react to different situations. In my opinion, this is clearly a sign of a problem with the PSU and/or the USB hubs themselves.

My question now is:
If I try connecting the mouse and keyboard to an external USB hub (preferably a powered hub), is there a chance the new USB hub will behave differently and/or be treated differently by the PSU?
Of course,  all this is in addition and independently of trying to connect all the devices using a power surge protected set of outlets.

All this seems to be guess work, as I don’t yet have those devices at hand to test it.

Anyway, it looks to me like replacing the PSU would be the most reasonable first step to take, except that I’m not able to do that alone. 🙁

After a 35 year career as a software developer, I now truly regret that I don’t have all the necessary hardware knowledge but have to always depend on a technicians’ help.

Thanks a lot & Best Regards to all !

Reply | Quote

iudith.m wrote:

So, to me, it’s clear the two different USB hubs (on the back side and the front side) are “treated” differently by the PSU and react to different situations. In my opinion, this is clearly a sign of a problem with the PSU and/or the USB hubs themselves.

My question now is:
If I try connecting the mouse and keyboard to an external USB hub (preferably a powered hub), is there a chance the new USB hub will behave differently and/or be treated differently by the PSU?
Of course, all this is in addition and independently of trying to connect all the devices using a power surge protected set of outlets.

Iudith, what you are experiencing could indeed be a faulty USB hub. I have a couple of PCs where a rear USB port has stopped working and I had to move the mouse to a different port. However, this could also be an early indicator of eventual catastrophic failure. One of the aforementioned PCs progressed over a couple of years’ time to shutting itself off at random, then to short-lived manual rebooting back into Windows, next to shutting down during POST, and finally it stopped working altogether. I’ve replaced every item in the PC case (PSU, cables, boot drive, RAM sticks, motherboard) and it remains dead as a doornail.

Regarding the use of an external USB hub, to my mind it’s a complicated enough situation that you’ll only find out for sure by trying it.

And as far as your technician insisting that it’s the fault of Windows 7 and refusing to explore the hardware side, may I suggest that it’s time to find a new technician.  🙂

Keep Running Windows 7 Safely for Years to Come  
Make Windows 10 look and work like Windows 7

Reply | Quote

You are so right regarding the technicians here. 🙁
The problem here is, in general, that no one is willing to investigate things at depth, they all prefer to just throw out and replace!
And yes, I’d be willing to do the same if I could simply take my hard disk with its OS on it and “transfer” it into a new machine but MS are criminals in that they don’t allow you to do this! Sorry, but I cannot find a milder word. :):)
One of the reasons I prefer to stick with Windows 7 is it’s no longer subject to all the failures of each OS update!!!
I still remember how we once had to reinstall the OS after a faulty Monthly update.

What you say is interesting, that your PC still did not work after you replaced everything.
So, what caused it not to work if everything was new? Except maybe the OS “did not like” the changes?

Back to my technician…
He does have a kind of “moral responsibility” to help me because this PC was assembled by him and I paid about twice as much for it as for my other PC, which is a name brand PC and has no such hardware problems.

If I had the help I expect, then I might even put aside a “last Windows 7 version” which could be used to reinstall the OS as a last resort. But he’s not willing to hear that.
All the good advice I followed, namely to backup my entire system, seems to be no help because I cannot use that backup to restore to a different/new hardware.

My feeling is that maybe all the internal USB connections were not exactly performing well from the start, and they got looser with the time. For example, after plugging the mouse in and out many times.
Also, the PSU is very noisy for a few minutes when the PC first starts and I told him about this once.

I am determined to make my tests using an external USB hub, but I will not give up trying to convince my technician to check the PSU and internal USB hubs.

I could choose to go to a computer lab, but the problem with those is the same as with anyone here. The first thing you have to do is to acknowledge they might have to reformat your hard disk… again, the easy way “to do things”… or always using a cannon to kill a fly.

I wish I had the knowledge myself and not be forced to accept what everyone likes to say based on the fact that “the customer does not understand anything.” However, with me, they have a harder time as I do understand the difference between hardware and software so they cannot tell me stories.

Reply | Quote

1. Is there any difference if the power cords for the various devices are swapped around to other wall receptacles in the room?
2. Can you run an extension cord to the computer from another room?

The light switch is described as a ‘regular old one’. At one time there were ‘quiet’ switches, which had a small, tilting, tube of mercury which closed the circuit.
3. Do you know ‘for sure’ that this switch indeed internally is only mechanical?
4. The room light: incandescent, fluorescent; single, multiple?
5. If incandescent, any change if the lamp is unscrewed and switch is still flipped? (Florescent fixtures typically are wired, not unpluggable.)
6. Are the light switch and the problem-computer outlet all on the same circuit?
That is, both are fed from the same circuit beaker? (Cf. Q2.)

You’ve perhaps already stated and I’ve missed it – keyboard and mouse.
Awkward to test.
7. Can you operate (with difficulty!) with only a keyboard or a mouse; and that from a back or a front USB port?
Perhaps both are needed at times even to get running, but can be unplugged during test scenarios.

Concerning your macro environment.
8. There is mention of ‘local’ considerations. Are you free to expand on that?
9. Is this in a house, or in a business, or a commercial area?
10. Are there any radiating devices or transmitters in the vicinity?
11. Is there a television in the room?
12. Is the computer near/against a wall, or is it in an open area?
13. – 99. Reserved for future use.
All these questions are to probes to see if there may be an RF and/or grounding issue.

1 user thanked author for this post.

Cybertooth

Reply | Quote

To test the regulation ability of the power supply to a lower input voltage, and incidentally (maybe) its ability to accommodate voltage dips and spikes –

Can you temporarily plug into the power strip/surge suppressor (if there is a shortage of receptacles you’ll have to unplug something else for a bit)
13. an extension cord with a lamp (desk lamp?), preferably ~100 watts or…
14. a high-wattage device such as a toaster, portable electric heater, or a hair dryer (but perhaps not on its highest setting)?
But check the nominal total watts of the testing load vs. the wattage (or current) rating of the strip/suppressor. Also verify the circuit breaker feeding the circuit can handle the entire load – all the wall receptacles on the breaker.
{watts = volts * amperes} Add ’em up.

Then cycle the lamp/toaster/whatever, on and off.

15. Is there another keyboard, or mouse, that you can swap in? A spare, or from another computer? (Checks if the problem is in the port or associated with the device.)

I agree the USB circuits are the most likely suspect, but other possibilities must be discounted first. This will help the technician too as he doesn’t have the time to run such exhaustive tests.

1 user thanked author for this post.

Cybertooth

Reply | Quote

Currently I don’t have any surge suppressor at my disposal, because the old ones I had seem to not work any more.
The green light for “protected” is off and, as far as I can remember, some long time ago it used to be on.  So hopefully the problem is not with the house electric wiring, but with those small devices themselves.

I already tried several mouses, keyboards, powercords, etc., with the same effect.

All the electric things which are unrelated to the computers I’d rather leave to a professional electrician.

I think the problem is already narrow enough for a technician to be able to put their finger on it after just a quick glance!

And even if a power surge protector could help “mask” the problem, I would still have to understand:
1. Why my other computer, connected to the same power outlet, does NOT have the same problem?
2. Why this computer only reacts to the on/off switch light if a mouse and keyboard are plugged into the front side USB’s and one or both of them are allowed to wake-up the computer?

Reply | Quote

This was my first thought but, unfortunately, I don’t have other wall outlets in the room and/or a long enough power cord to test it in another room.
Well, except if I disconnect the PC from the router and physically carry it to another room just for testing… not so easy for me either.

The light in the room is fluorescent, a single tube.

There is no television in the room.

The mouse/keyboard issue relates to whether they are allowed to wake-up the computer or not.
If only one of them is allowed, then the problem gets triggered.
If none of them are allowed, the problem goes away.

The huge difference between the back/front side USB’s is that, with the back side USB’s, the wake-ups were totally random. With the front side, they are deterministically related to the the switch light on/off.

I don’t exactly understand where the USB’s are receiving their power from when the PC is in sleep mode.
Anyway, it looks like the USB is powered differently if a mouse/keyboard is allowed to wake-up the computer… which to me looks “too wise” for a PSU which is a “dumb” device.

I’m not a USB specialist but, as far as I understood, an external USB hub with a power supply is supposed to be powered from its own power supply and not from the PC.
In that case, how will the mouse/keyboard behave when connected to such an external USB hub? Will all the “wake-up knowledge” be carried on through the USB port to which the external HUB is connected?
What if the specific external USB ports are powered off (which cannot be done with an internal USB port)?
I will definitely test this.

Regarding my “local” considerations:
What I wanted to say is that all the electrical devices should be compatible with the local electrical installations standards, which are different from country to country.
For example, the outlet voltage and the electrical cord plugs are different here than in the US, so I need to rely on local stores for acquiring anything.

Reply | Quote

Thank you; you’ve cleared several thoughts.
5 revisited. Can you remove the fluorescent tube from the fixture and flip the switch?

Reply | Quote

I hope you have luck getting someone to test the issue (wiring in the case or worn out USB ports could be related), but before doing that, I think I would hire someone specifically to install a new power supply.  They are as you mention reasonably priced.  If the manual for the model of motherboard is posted on a website most repair should be able to install one fairly straightforwardly at a fixed price for an amount of work they know ahead of time.

Reply | Quote

Hello Judith.m. You said, “Currently I don’t have any surge suppressor at my disposal, because the old ones I had seem to not work any more. The green light for “protected” is off and…”.

I think you may have your answer right there. The job of a surge protector is to harness (called clamping) the surge and run it off to ground.

PaulK and AleJr have some very good ideas but, before people go further with this issue, you REALLY need to get a decent surge protector with “noise filtering”. These are not the cheap ones for $10 but the better ones that do cost a bit. I currently have a good quality surge protector (with noise filtering) on my equipment and have had it for years. The “Grounding and “Protection” light are still on to let the consumer know all is well. I have everything for that PC plugged into the Surge Protector.

NOTE, do not plug a Laser Printer into a UPS (Uninterruptible Power Supply) unless it is the sole protector for that printer and is -capable- of the high current needed for Laser printing.

You do have good grounded outlets right? If not, the more expensive surge protectors will have a “fault ground or GND Fault” light turn on to warn you.

Your other statement, “Why my other computer, connected to the same power outlet, does NOT have the same problem?” could very well point to the Power Supply being a bit sensitive, or something on the Motherboard is sensing a voltage change.

You also said, “I don’t exactly understand where the USB’s are receiving their power from when the PC is in sleep mode.” It’s normal for PCs to supply power to a USB device while sleeping. Otherwise, how would it know you moved a USB mouse or tapped the USB keyboard? Are you using a USB mouse and keyboard or PS2?

I would take the suggestions of others here and do a few tests, but feel you do need to invest in a good surge protector to protect the equipment you currently have.

Good Luck.

Reply | Quote

Yes, I do intend to buy a surge protector but there are not too many types available here and, due to plug compatibility issues, I can only choose one of those.

The main difference is there are units with one single outlet, or boxes (sets) with several outlets so the protection works for the entire box. I cannot really say how good they are until I try them.

I suppose the grounding of the outlets in the house is OK, but I also cannot know this until I try to use a protector or call an electrician to check them.

Thanks a lot. 🙂

Reply | Quote

You “should” be able to get an outlet tester at the same place where you’d buy surge protectors and power cords.

It’d look somewhat similar to this but with prongs that match what your particular wall outlets use.

 

The indicators, at the bottom of this particular tester, light up in the various different patterns shown to indicate whether an outlet is wired correctly or not. I’m not sure what the price would be where you’re located, but here in the US they’re typically less than $15.

Of course, if it indicates something’s not wired correctly, you need to have a licensed electrician fix it.

1 user thanked author for this post.

Cybertooth

Reply | Quote

PMFJI, and this may be somewhat off-topic, but recently BOTH my computer and my wife’s computer (both home-built by me with Gigabyte mobos) would almost immediately (~1 min.) waken spontaneously from sleep. Event Viewer showed they were waking up to:

AMD USB 3.10 eXtensible Host Controller – 1.10 (Microsoft).

which led me down a merry rabbit hole for days, until I found a reddit post suggesting I analyze wake alarms (WakeOnLAN is off in the BIOS/UEFI) by running a cmd prompt:

powercfg -devicequery wake_armed

which showed 3 devices were reported: mouse, keyboard and something called the Realtek Gaming 2.5GbE Family controller

I unchecked “allow this device to wake computer” on all of these, and the problem stopped. I next sequentially re-enabled the mouse and the keyboard, but the problem did not reoccur.

Don’t know if this will help iudith, but it MIGHT help the OP (who now seems to be missing).

HTH.

Zig

1 user thanked author for this post.

Cybertooth

Reply | Quote

iudith.m wrote:

I don’t exactly understand where the USB’s are receiving their power from when the PC is in sleep mode.

The motherboard controls the power for the USB ports not the PSU!

Also, there’ll be some settings in the BIOS that control whether they receive power or not and whether they can “wake” the PC when the it’s in sleep mode. Note, the BIOS settings will override the one’s in the Windows OS (i.e. if they’re set to not allow the PC to be woke up in the BIOS, the settings in Windows that control that function won’t have any effect.)

iudith.m wrote:

The light in the room is fluorescent, a single tube.

This is the most likely cause of your problem!

Fluorescent tubes are notorious for emitting high levels of EMI/RFI (Electromagnetic Interference/Radio Frequency Interference) when powered on and, what’s probably happening, is there’s enough EMI/RFI bleed thru into your keyboard/mouse to make the PC think they’re “in use” and wake it up.

The problem is “random” when using the “rear” USB ports because the metal of the PC case is between the USB ports and the light which blocks most (but not all) of the EMI/RFI so… sometimes it’s strong enough to cause the problem, other times it’s not.

When you use the “front” USB ports, there’s likely little to no metallic objects between them and the light to block the EMI/RFI so the effect happens every time you flip the light switch.

It’s not affecting your “other” PC because it’s either far enough away from the light that the EMI/RFI is too weak to effect the mouse/keyboard or there are enough metallic objects between it and the light to block the EMI/RFI being emitted.

To test for possible EMI/RFI, place something metallic between the light and the USB ports being used by the keyboard/mouse (a small sheet of aluminum foil would be perfect) to block any EMI/RFI and then turn the light on/off.

If the wake-up doesn’t happen, you’ve found the culprit and will have to figure out a more permanent solution to block the EMI/RFI (which “might” be as simple as replacing the “ballast driver” for the fluorescent tube.)

If the wake-up still happens, then at least you’ve eliminated possible EMI/RFI from the fluorescent light as being the cause.

2 users thanked author for this post.

PaulK, satrow

Reply | Quote

This information is very interesting!
Anyway:
All 3 of my computers (2 desktop + 1 laptop) are near one other, on the same table, in the same room. However, only this one specific desktop PC is affected by turning the light switch on/off.

When the keyboard and mouse were connected to the rear USB’s, I experienced completely random wake-ups, without turning the light switch on/off and even without being present in the room or in the house at all! This is why I tried to move the keyboard/mouse to the front side USB’s.

The laptop has a mouse connected to a side USB and it does NOT react to the light switch!

If the light switch affects the USB ports and NOT the keyboard/mouse, then maybe the solution of using an external USB hub could work, because as far as I understand, those hubs are made of aluminum, so are kind of similar to the rear and side metallic computer case.

It would be hard to isolate anything from the light, because the fluorescent tube is of course located on the room ceiling, so the light spreads everywhere! I think maybe I’ll try using a piece of kitchen baking aluminum foil and cover the USB ports with it.

I was sure that the light switch only “communicates” with the USB ports through the electric wiring so I’m now on my way to try to use a long power cord extender to plug the PC power cord into an outlet in a different room… just to see what happens.

What exactly is the “ballast driver” of a fluorescent tube and how can such a thing be replaced?

Really interesting info, thanks a lot. 🙂

Reply | Quote

The ballast will be in the fluorescent tube housing, typically right behind the tube itself, and look something like this.

Here’s a video showing How to replace a fluorescent lamp ballast.

Reply | Quote

iudith.m wrote:

All 3 of my computers (2 desktop + 1 laptop) are near one other, on the same table, in the same room. However, only this one specific desktop PC is affected by turning the light switch on/off.

Try swapping the keyboard/mouse combo used on the PC that has the problem with the ones used on the PC that doesn’t have the problem?

Did the problem “follow” the keyboard/mouse combo (i.e. the other PC now wakes when you turn the light on/off) or does it still only happen on the original PC with a different keyboard/mouse combo plugged in?

If the problem moved to the other PC, then there’s something wrong with “that particular keyboard/mouse combo” and the best solution would be to simply replace them.

If it doesn’t move, then there’s something wrong with the original PC.

BTW, the above advice only applies if your testing with a longer extension cord and/or the aluminum foil eliminates possible wiring/fluorescent light issues as a possible cause.

Reply | Quote

iudith: Please clarify. Are the keyboard and mouse separate devices, each with its own cable-and-plug? Or is is there only one USB cable into the computer for both?

Have you tried just removing the fluorescent tube? It may not help, but it is one less variable to be considered.

Extension cord from another room: You’ve verified that that outlet is on another breaker circuit? We’ve been assuming that the power for the computers’ complex and the power for the overhead light are both on the same circuit.

alejr: A ‘blocking’ foil. Wouldn’t this need to be grounded/earthed to the computer chassis/frame? Mice and KBs often are plastic. If ungrounded/unearthed, isn’t it possible for any RF to be reflected off of something – and thus be inconclusive? I’m still wondering about some grounding/shielding aberration, which could permit some kind of resonance.

Reply | Quote

The mouse and keyboard are separate USB plugs.

I have tried to cover completely the USB ports with aluminum foil, sticking it to the computer case, and at the beginning I had the impression that it helped but, afterwards, it still did not.

I cannot remove the fluorescent tube… really not!
But, again, if this is a culprit indeed, then I still cannot understand how come that ONLY this single USB hub is affected by switching light on/off ?!?

Now finally I got a very long power cord extender from a friend, and plugged the computer in another room, at many meters distance… a totally unrealistic solution, but I still want to test the behavior.

So, now I have both the aluminum foil still covering the USB ports and also the PC plugged to a power outlet far awa.
If this will work, then I will remove the aluminum foil to see indeed what is the problem.

Again, in my opinion, the USB is flawed, because a USB is supposed to distinguish between a keyboard key being pressed and any other effect!

If instead, the culprit is the power line, then I guess the faulty effect should be similar to pressing the power button, and NOT similar to pressing a keyboard key.
That is, in this case, I suppose the computer would wake up regardless of whether a mouse/keyboard are plugged in at all… however, this DOES NOT happen!

1 user thanked author for this post.

Cybertooth

Reply | Quote

PaulK wrote:

A ‘blocking’ foil. Wouldn’t this need to be grounded/earthed to the computer chassis/frame? Mice and KBs often are plastic. If ungrounded/unearthed, isn’t it possible for any RF to be reflected off of something – and thus be inconclusive?

Any object placed between something emitting EMI/RFI and another object will block/reflect some of the EMI/RFI being emitted; even if it’s not grounded.

“Metallic” objects like aluminum work best because they “reflect” almost all of the EMI/RFI away from the direct “source to destination” path (which is what we want in this instance.) In fact, if it’s curved correctly (like an antenna disk), a metallic object can actually “reflect” the incoming EMI/RFI down to a single focal point which concentrates and amplifies it making the effects much more pronounced.

EMI/RFI is a form of electromagnetic radiation (magnetic fields, radio waves, microwaves, infrared, visible light, ultraviolet, X-rays, gamma rays, etc.) and follows the inverse-square law of intensity (i.e. power) vs distance. That means, even if it gets “reflected” off multiple surfaces to still reach the destination, the “distance” it has to travel to get there increases and, as shown in the below image, the strength drops off very quickly the further it travels.

The inverse-square ratio means:

• 2 x distance = 4 x less power
• 3 x distance = 9 x less power
• 4 x distance = 16 x less power
• etc., etc.

So any “reflected” EMI/RFI will be must less likely to cause the same effect.

There’s also much less chance a “reflected” EMI/RFI signal will actually reach the 2nd object than if there’s nothing blocking the direct path between it and the EMI/RFI source.

For more info see How EMI/RFI shielding manages interference.

2 users thanked author for this post.

Cybertooth, PaulK

Reply | Quote