Windows 10 Recurring Crash to Black Screen

[ElPengwino]

Symptoms:


I have crash that occurs pretty frequently on my machine. When it occurs the screen goes almost completely black, filling with very dark "snow" (noise) that upon closer inspection is actually tessellated. If any audio was playing at the time of the crash, then it is replaced by the "robot growl" (a very rough, constant noise), otherwise no sound plays after a crash. The crash can only be resolved by performing a hard reboot.


System Specs:


CPU: Intel Core i7 4790K

Motherboard: MSI Z97 Gaming 5

Video Card: AMD Radeon R9 390, model 100382NTOC-2L (driver version 26.20)

RAM: G.SKILL 2x8 GB DDR3 1600 SDRAM

PSU: EVGA 850 W

OS: Windows 10 x64 (version 1803)


Attempts to fix/diagnose problem:

• I ran mdsched to check the RAM, but it did not find any problems.
• I tried to repair any corrupted system files by running the following commands:
  • DISM.exe /Online /Cleanup-image /Restorehealth
  • sfc /scannow
• I consulted the Reliability Monitor to try to find the cause of the crash.
  • I can see events that seem to have been logged after I started the machine back up after a crash that only say "Windows was not properly shut down. The previous system shutdown was unexpected."
  • The Reliability Monitor shows a "Hardware Error" just before the first such crash. The problem event name is listed as a "LiveKernelEvent" with code 141.
    • I am very inexperienced with working with the Event Viewer, and have been unable to isolate this event there to try to glean further information. Additionally, the Event Viewer in my current version of Windows seems to crash when the user attempts to make a custom filter.
    • Searching online for "LiveKernelEvent code 141" seems to indicate that this event can stem from a wide variety of issues such as failing hardware, corrupted drivers, etc as opposed to a specific cause.
• I have played games with the case open, but the crash still occurred.
• I carefully vacuumed as much dust as I could out of the case and changed the video card to a different slot with the same interface. The crash still occurred.
• I ran FurMark for 20 minutes, with GPU temperatures staying between 80 and 85 degrees C for almost the whole test. The computer did not crash.
• I ran Prime95 for 15 minutes, with all CPU cores staying at 100% utilization for the whole test and staying between 95 and 100 degrees C for almost the whole test. The computer did not crash.

Additional Observations:

• The crash is very random with no easy way to quickly and reliably reproduce it.
• Three things make the crash more likely (but not guaranteed) to occur:
  • If a crash has occurred recently then it is very likely that a crash will occur again. If I reboot the computer when a crash occurs and continue using it immediately after the reboot it will almost always crash again in under 5 minutes.
  • Gaming seems to induce the crash more reliably than just surfing the Web. However, I have had the machine crash with only simple programs open like the Reliability Monitor and Event Viewer.
  • Switching between programs by using alt-tab (especially if the current application is running in fullscreen) has caused the crash.

Thank you for any assistance!

Continue reading...