Question Random Black Screens - Forced to Reboot

Jan 8, 2021
6
0
10
0
My PC has become unstable and locks up completely every 5-10 minutes once I'm in Windows, no matter what I'm doing (browsing, working on office docs, etc.) Not even gaming or streaming. The symptom is this: the screen just goes completely black, and no keystrokes do anything (ctrl + alt delete, ctrl + shift + win + B, win key, etc.) Also no amount of moving the mouse around and clicking does anything.

The LEDs on the PC are still lit when it happens, and the fans are still turning like normal. The only way to recover functionality when this happens, is to hold in the power button on my PC and power cycle the system. Once I do that it works fine on booting but once I sign into Windows it will not last but 10-15 minutes until it happens again.

(This has progressively gotten worse before reaching this point. It happened only on occasion like 1-2 times a week starting about a month ago, and over the last 2 or so weeks has finally gotten worse to the current situation.. happening multiple times daily to the point it is now: happens within minutes of logging in every time.)

I have confirmed If I boot Windows 10 in SAFE MODE, my system is stable. I can use it for hours without having the issue happen. In fact, I am on the very same PC in safe mode writing this post right now, and it's been online for 4 hours so far with no crash (this would be unheard of if I wasn't in safe mode.)

Here are some troubleshooting details:

- When it first started happening, I would get amdkmdap stopped responding in Event Viewer, along with kernel-power warning (presumably from when I power cycle the system?) I followed some threads including several posted on these forums, and after trying out clean GPU Driver Installs using DDR, I no longer get that event in event viewer any longer, though the problem does still happen. When the problem happens now, there's nothing really in event viewer other than kernel-power.

- I have tried to reboot into safe mode, run DDU, and then re-install the drivers fresh after rebooting. When I'm in safe mode I'm good and stable. After the drivers are installed, the problem comes back. I tried the December 2020 and November 2020 driver version, as well as the one Windows Updates places that are from 2019. Same problem no matter the version. So that's 3 different versions of the AMD drivers with the same issue (seems like hardware??)

- I ran the native Memory Checker and Disk Checker tshooting tools in Windows 10. Neither reported any issues.

- I have not done any hardware layer tshooting yet inside the case. Have not tried reseating anything, etc. (Does it make sense to do this if Safe Modes protects from the issue?)

Here are my system stats

OS Edition: Windows 10 Home
Version 20H2 (OS Build 19042.685)

GPU: RX-570 4GB
CPU: Ryzen 5 2600 (so no integrated display)
MOBO: B450M DS3H AM4
PSU: CoolerMaster 650 Watt 80 Plus
RAM: Vegence LPX DDR4-2400 16GB Kit (2x 8GB Sticks)

What is my next steps for troubleshooting? I had the idea of trying to run outside of safe mode with default Microsoft Display Drivers to see if I remain stable, unfortunately Windows Updates is forcing the driver to re-install every time, no matter if I turn Device Installation settings off or not.

With how often this started happening it feels like a hardware component, is dying... but it remains perfectly stable in Safe Mode so now I'm actually not so sure anymore? Could it still be a hardware issue, or does Safe Mode indicate it is more of a software/driver issue?

Any advice on how to go from here to help isolate the root cause is greatly appreciated!
 
Jan 8, 2021
6
0
10
0
Here is an update on what I have tried and what the results are

I have updated to the latest B450 Chipset Drivers listed at amd.com

I have flashed the newest version of BIOS Available for my B450M DS3H Motherboard at Gigabyte's web site.

I have reset the BIOS to "optimal default settings" (I never changed anything to begin with, anyway.)

After doing all this I booted into Safe Mode, ran DDR to clear the drivers out, and then re-installed fresh drivers from AMD.Com.

Problem is NOT solved, but the behavior is changed. Now instead of just locking up and becoming unresponsive when I black screen, now the PC reboots itself, and in Event Viewer I get a bugcheck 0x00000116 message.

When you look that bugcheck up, it says that the display driver timed out, and did not recover. So, that basically narrows this down to the GPU in my opinion.

I'm still stable if I run in Safe Mode. And today I discovered I'm stable if I run outside of Safe Mode with Microsoft Default Display Adapter.

That makes me wonder, is it really the GPU HARDWARE that is bad and needs to be replaced? Then why is my system just fine and stable when no drivers are installed for it? However, no matter which different version of the driver I install, it seems the problem is coming back every single time.

If anyone could please give me some advice for next steps, I'm at my wits ends. I'm planning to just buy a cheapo graphics card from Best Buy and just swap it out and see...
 

ASK THE COMMUNITY