Windows 10: Black Screen Of Death

DWD1961

[H]ard|Gawd
Joined
Nov 30, 2019
Messages
1,314
UPDATE:

I unplugged all of my USB peripherals in order to let the drivers re-establish. Then I uninstalled the AMD chipset USB drivers, which caused the system to crash and auto-reboot. I've slowly been plugging back in my USB monitor connection, hubs, and other devices. So far no black screen at all. If it holds, I think it was a corrupted USB driver. I have no idea which one, but I have a suspicion it was the Gigabyte Sidekick one screen OSD that requires a USB connection. After plugging the monitor, the Sidkick program would no longer start, giving me a "Check USB Connection" error. I uninstalled and reinstalled the software, and it worked. Anyway, that's my best guess.

ORIGINAL POST:



I've been trying to trouble shoot this problem for about three weeks now and I'm having no luck.

System Info:
Windows 10 Pro Latest
Version 10.0.19044 Build 19044
Gigabyte B450 I AORUS PRO WIFI
AMD Ryzen 5 3600
Gskill RAM 16.0 GB
---------
  1. This happens randomly about average 2 times a day. Seems to be getting worse.
  2. It has never happened on start up.
  3. Screen will freeze, then black screen and "no HDMI input"
  4. All peripherals unresponsive
  5. Computer music remains playing and I can talk on discord, etc. Just a black screen with no way to interface. It's like the video card just went away.
  6. Requires hard restart.
  7. No flags in Device manager.
I was thinking it was driver related, but I haven't used the Windows basic driver yet.

Or, the card itself, but I only have a really old GTX295 card to test it, and I'm pretty sure that card isn't even supported anymore. And, it may not be viable itself since it sat in a garage for over 6 years. I guess I could buy a cheap (by today's standards) card just for testing purposes.

Since this card only has one HDMI output, and two mini DP outputs, I was thinking about getting a mini DP to HDMI converter or just a mini DP cable and try that (I've already tried a different HDMI cable).

Yes, I reseated the graphics card. Funny as it sounds, started the system and within 5 minutes, black screen.
 

Attachments

  • Capture.PNG
    Capture.PNG
    48.6 KB · Views: 0
  • Capture2.PNG
    Capture2.PNG
    11.3 KB · Views: 0
Last edited:
what card and what have you tried? ^^ thats one place to start that does involve much.
 
try bumping the power in afterburner, see if it helps. my old 280x needed 10%+ or id get glitches, worth a shot.
'm Not using the 295, only the 7950. Also, the card hasn't been showing any signs of malfunction until recently. You would think if it were starving for powerr, it would ahve shown up long before several weeks ago.
 
'm Not using the 295, only the 7950. Also, the card hasn't been showing any signs of malfunction until recently. You would think if it were starving for powerr, it would ahve shown up long before several weeks ago.
ok, didnt say anything about the 295...
yeah, it probably means its dying. try it or not, your call. gl.
 
At any time. Idle or gaming.
Its time to inspect for unseen damage.
a) carefully inspect PSU connectors, 12V for CPU along 24 pin large, check with magnifying glass for pins that appear half burned.
b) Test the system with another PSU.

And here ends the remote assistance.
 
I think it may have something to do with USB drivers. I haven't been able to nail it down, but I have reduced it by keeping all unessential USB devices unplugged.
 
Search bios for OC fabric and memory switch if there turn it off whether you oc them with pbo or not. I had a b450 mini board and that switch caused some wacky usb issues until turned off and I still could run memory and fabric clock at 1800. Can't remeber which board it was either the aorus or the tomahawk as it's been awhile back and I had both.
 
black screen of death is a somewhat common amd video card/driver issue i have had a few times in the past and ended up fixing them by doing a full ddu clean and clean installed latest working driver for the card. has zero to do with usb in any way
 
black screen of death is a somewhat common amd video card/driver issue i have had a few times in the past and ended up fixing them by doing a full ddu clean and clean installed latest working driver for the card. has zero to do with usb in any way
Well, god damn then. I'll download DDU and get 'er done! Thanks!
 
black screen of death is a somewhat common amd video card/driver issue i have had a few times in the past and ended up fixing them by doing a full ddu clean and clean installed latest working driver for the card. has zero to do with usb in any way
It actually did it again after cleaning and installing. This is an end of life card (AMD 7950), and the last drivers AMD published last summer were the cards "end of life" drivers. I have a suspicion they have a 7950 bug in that last driver update. I roiled back to the driver to just before that, which was 4/30/22. So far, no problems.
 
It actually did it again after cleaning and installing. This is an end of life card (AMD 7950), and the last drivers AMD published last summer were the cards "end of life" drivers. I have a suspicion they have a 7950 bug in that last driver update. I roiled back to the driver to just before that, which was 4/30/22. So far, no problems.
very possible
 
very possible
Nope! Still happening. It seems like the longer my rig is up, the more chance it has to BSOD-ing. Tongiht it not only BSODed but nothing worked. The rig was still running, but everything else was down, music, Discord, etc., not just the mouse and keyboard. Going to check my system for stability tomorrow. See if it tosses any errors.
 
Nope! Still happening. It seems like the longer my rig is up, the more chance it has to BSOD-ing. Tongiht it not only BSODed but nothing worked. The rig was still running, but everything else was down, music, Discord, etc., not just the mouse and keyboard. Going to check my system for stability tomorrow. See if it tosses any errors.
Definitely check system logs for solid clues
 
Definitely check system logs for solid clues
Checked sys logs (again) and this time found what I think is the problem. It was USB related, but not a problem with USB.

"User-mode Driver problems. The device HID-compliant headset (location (unknown)) is offline due to a user-mode driver crash. Windows will attempt to restart the device 4 more times. Please contact the device manufacturer for more information about this problem."

This is a headset plugged into the USB port. That's why when I unplugged it, no crashes. Or, when I wasn't using it. At least that's what I think for now. Checking all of the Critical logs since 11-21, when I got the headset, the only critical errors are created from the headset driver crash. Then I get the unknown driver error and tehn finally Kernal crash, since I have to do a hard reset.
 

Attachments

  • PolyDriverIssues.txt
    1.5 KB · Views: 1
  • DriverPolyHeadset.PNG
    DriverPolyHeadset.PNG
    6.8 KB · Views: 1
Have you checked your RAM using Memtest? I would test 1 stick at a time to rule that out. https://www.memtest86.com/
No. I'll test all of the hardware, though. I still think it is driver related. Many others on the internet have psoted about it.

Description
The Windows User-Mode Driver Framework detected that a driver host-process did not complete a critical operation within the allowed timeout period.
This report contains information about the process and the drivers running within and will be used to improve the quality of these drivers.

It's not the Poly headset. It's not a Poly driver. It's a Windows provided driver. It keep crashing. Then it takes the entire system with it. well, not really. Most of the time it will just take the graphics input out and all USB connections, but music is still playing, etc.

But the crash is ALWYS casued by: "User-mode Driver problems. The device HID-compliant headset (location (unknown)) is offline due to a user-mode driver crash. Windows will attempt to restart the device 4 more times. Please contact the device manufacturer for more information about this problem."

Problem signature
Problem Event Name: WUDFHostProblem
EventClass: HostProblem
Problem: HostTimeout
DetectedBy: 2
UMDFVersion: 2.31.0
ExitCode: 103
Operation: 3
Message: 11b00
Status: ffffffffffffffff
HardwareId: HID\VID_047F&PID_02E6&REV_0315&MI_03&Col05 <--- That the headset driver in Windows.
OS Version: 10.0.19044.2.0.0.256.48
Locale ID: 1033

Extra information about the problem
Bucket ID: 35db347156490ede2d5120d3adfb61d5 (2112505793742594517)

I tested it last night with a Wireless (not Bluetooth) headset, and got the same error. If you search "The device HID-compliant headset (location (unknown)) is offline due to a user-mode driver crash, " you will see many, many problems with this little nasty bug. I don't think they are all related to RAM or hardware failure.

There were so far 3 solutions thta may or may not work:

1. CPU power cable was loose
2. Revoking Windows rights to update HID headset driver. For some reason, that fixed it for that specific person.
3. USB selective suspend setting > Disable (mine always was, I think, but last after installing new chipset drivers, the power plan was on Ryzen balanced, and the suspect option was "enabled." I disabled it.

It's really fucking pissing me off. Sick of it.

Most report reinstalling Windows doesn't solve it ether.
 
Last edited:
PASSED MEMTEST AND PRIME95 Blend Mode:

I really have no idea what to do. It's so random there isn't anyway to nail it down. The last time it crashed, I looked at windows logs and it said there was a hardware error. Before that the error was "HID headset." Well, I disabled the HID Headst from loading, and then got the "hardware error."

I ran metest386 and Prime95 on blend. The computer crashed again when Prime95 was running, but it ran for 1.5 hours and the logs said there were no errors (rounding errors).
Then I ran Memtest 383 and it ran it's entire test (4 hours) and gave a pass.

When it crashes, it will sometimes stutter, then keyboard input and mouse input freezes. It takes out the USB stack because I have no keyboard or mouse control. Then it shuts the HDMI link to the screen down, and I get the "HDMI no signal" from the monitor.

I'm going to uninstall the AMD drivers. I already rolled them back to 4/20/2021 which was before I started having problems. I'll let Windows load default drivers for it and see what happens.

If that doesn't work, I'm going to load Win 11 fresh.

I also have a really old nVidia 295 card, and I mean really old, like back in 2008. It's was sitting in a garage for 6 years, and now has sitting on an indoor shelf for 3 years, so I don't know if it would be a good test or not. I guess I could try it.

Any other ideas?

Last three crash logs:

Description
A problem with your hardware caused Windows to stop working correctly.

Problem signature
Problem Event Name: LiveKernelEvent
Code: 141
Parameter 1: ffffe70f6d90b460
Parameter 2: fffff8008ed32160
Parameter 3: 0
Parameter 4: 348
OS version: 10_0_19044
Service Pack: 0_0
Product: 256_1
OS Version: 10.0.19044.2.0.0.256.48
Locale ID: 1033

Description
A problem with your hardware caused Windows to stop working correctly.

Problem signature
Problem Event Name: LiveKernelEvent
Code: 141
Parameter 1: ffffe70f6bbea010
Parameter 2: fffff8008ed32160
Parameter 3: 0
Parameter 4: 4
OS version: 10_0_19044
Service Pack: 0_0
Product: 256_1
OS Version: 10.0.19044.2.0.0.256.48
Locale ID: 1033

Description
A problem with your hardware caused Windows to stop working correctly.

Problem signature
Problem Event Name: LiveKernelEvent
Code: 141
Parameter 1: ffffe70f6daf7430
Parameter 2: fffff8008ed32160
Parameter 3: 0
Parameter 4: 4
OS version: 10_0_19044
Service Pack: 0_0
Product: 256_1
OS Version: 10.0.19044.2.0.0.256.48
Locale ID: 1033

Then there was one of these with a different code 1cc:

Description
A problem with your hardware caused Windows to stop working correctly.

Problem signature
Problem Event Name: LiveKernelEvent
Code: 1cc
Parameter 1: ffffe70f68547290
Parameter 2: ffffe70f6c0f7080
Parameter 3: 362c6
Parameter 4: 96
OS version: 10_0_19044
Service Pack: 0_0
Product: 256_1
OS Version: 10.0.19044.2.0.0.256.48
Locale ID: 1033

But most of them are 141 kernel events.
 
Last edited:
Wow what a PITA! I would pull the mobo out of the case mount it on some cardboard to rule out a possible short. Try a different power supply.
 
Wow what a PITA! I would pull the mobo out of the case mount it on some cardboard to rule out a possible short. Try a different power supply.
something on that brand new motherboard is clearly not fully stable.....the windows stopped working is way more than just black screens like i was getting from my amd card driver's
 
Try the headset in a different port. If it works in the new port without crashing, there's probably something messed up in windows' registry (a separate entry is created for each device, for each port it is used in).
 
The same day of my last post, that night I let Windows run a chkdsk.
chkdsk C: /f /r /x

I couldn't sit there and wait for it to reboot, but when I came back it said "fixing." I haven't had a black screen since that time (about 11Am 1/24). After shutting down my rig last night, and then starting it this morning, it got stuck on the spinning dots. I restarted it again, and it booted right into Windows. Since I ran chkdsk, I haven't had one crash. I don't know what could cause the black screen of death from the drive or what was wrong with the drive that could have caused those problems.

However, after approximately 36 hours, not one crash. I've let the system run continuously today since 8:30AM. God damn I'm glad this is over (hopefully).
 
Try the headset in a different port. If it works in the new port without crashing, there's probably something messed up in windows' registry (a separate entry is created for each device, for each port it is used in).
First thing I did way back.
 
Well. Literally 42 minutes after posting:

Description
A problem with your hardware caused Windows to stop working correctly.

Problem signature
Problem Event Name: LiveKernelEvent
Code: 141
Parameter 1: ffffd40373cf2010
Parameter 2: fffff8027ccd2160
Parameter 3: 0
Parameter 4: 34c
OS version: 10_0_19044
Service Pack: 0_0
Product: 256_1
OS Version: 10.0.19044.2.0.0.256.48
Locale ID: 1033
--

This time no black screen, just everything locked up, music still playing in the background. I'll reload Windows, then after that, just replace the video card when I can afford it, and just live with it. Hopefully it's a Windows update bug of some sort. Otherwise, I'll just have to start randomly replacing all of my hardware until the problem stop or I have a completely new rig.
 
I've been waiting on Windows 11 to be ready for my rigs' install, and today it was there. I didn;t want to fresh install 10 just to have to reinstall 11 again.
 
Read up on that error and it seems it is likely the GPU on its way out.
How did you find that information? It is the most likely hardware failure, since this card is 10 years old, and for 6 of those years, it was in an outside garage, not a controlled environment. The capacitors got cold and hot over and over again, although slowly with the weather change.
 
How did you find that information? It is the most likely hardware failure, since this card is 10 years old, and for 6 of those years, it was in an outside garage, not a controlled environment. The capacitors got cold and hot over and over again, although slowly with the weather change.
I just googled it and it seems a bunch of people who had it eventually tracked it down to their GPU. Obviously I'm not 100% certain or anything it just seems the most likely candidate.
 
How did you find that information? It is the most likely hardware failure, since this card is 10 years old, and for 6 of those years, it was in an outside garage, not a controlled environment. The capacitors got cold and hot over and over again, although slowly with the weather change.
swap the gpu with the ancient one you have and report back....either it fixes it or it doesnt....IF it doesnt then you have an unstable new setup that can possibly be fixed with bios settings or finding the hardware thats causing it. windows has bugs but nothing usually that bad...its even possible to be a bad storage device....time to swap parts till the problem is found
 
I just googled it and it seems a bunch of people who had it eventually tracked it down to their GPU. Obviously I'm not 100% certain or anything it just seems the most likely candidate.
Mostly I see an unsolved mystery for LiveKernelEvebt 141 and Location 1033. It seems no one has solved this problem, and reported back that it is actually solved. Once you get it, you just start replacing hardware until it goes away, and that last piece of hardware isn't necessarily the bad part either. It's a complex combination of many things, is my assumption, that line up just perfectly with your hardware and software. Some people have reported having it for over two years with no solution, even after RMAing their video cards and replacing power supplies.

If I wanted to spend the next 100 hours with a spread sheet and logical block analysis, I could find it. But, that's not worth the time to me. I'll just live with it until I can start replacing hardware when I need it. I need a card now, can't afford it, but that will be my first replacement. Then I'll upgrade my CPU, then MB, then RAM, and if the problem persists, the power supply. I have 3 NVME SSDs I use for backup, so I can just swap those for testing.
 
swap the gpu with the ancient one you have and report back....either it fixes it or it doesnt....IF it doesnt then you have an unstable new setup that can possibly be fixed with bios settings or finding the hardware thats causing it. windows has bugs but nothing usually that bad...its even possible to be a bad storage device....time to swap parts till the problem is found
Yeah I'll try it, unless it's too big for my current build. Those old GTX 295s were huge fucking pigs.
 
Mostly I see an unsolved mystery for LiveKernelEvebt 141 and Location 1033. It seems no one has solved this problem, and reported back that it is actually solved. Once you get it, you just start replacing hardware until it goes away, and that last piece of hardware isn't necessarily the bad part either. It's a complex combination of many things, is my assumption, that line up just perfectly with your hardware and software. Some people have reported having it for over two years with no solution, even after RMAing their video cards and replacing power supplies.

If I wanted to spend the next 100 hours with a spread sheet and logical block analysis, I could find it. But, that's not worth the time to me. I'll just live with it until I can start replacing hardware when I need it. I need a card now, can't afford it, but that will be my first replacement. Then I'll upgrade my CPU, then MB, then RAM, and if the problem persists, the power supply. I have 3 NVME SSDs I use for backup, so I can just swap those for testing.
If they had the problem for 2 years they haven't really put forth the effort in figuring out what the actual problem is. It can be difficult if you don't have a full set of spares, which is why most don't bother, they just pick a part and RMA it.
 
Back
Top