Complete Freeze - Kernel-Power, Event ID 41, Task 63

Hey,

I have a machine I build myself over a year ago that has slowly descended into the point where it's unusable. It started where it would freeze every once in a while, no BSOD, had to hard reset. Once every other week, it was no biggie. Now, it won't go past 20 minutes without freezing up. I know this forum says BSOD only but I'm at wits end.

The event log basically says nothing. Kernel Power Error, Event ID 41, Task 63. Details show this:

BugcheckCode 0
BugcheckParameter1 0x0
BugcheckParameter2 0x0
BugcheckParameter3 0x0
BugcheckParameter4 0x0
SleepInProgress false
PowerButtonTimestamp 0

I've tried:
- Audio drivers updates
- Multiple power management settings (not sure that it matters as it's a desktop)
- Memtest
- Reseating memory and video card
- Swapped SSD drive (old drive reviews had users having similar issues)
- Rollback to older windows 10 versions (and stopped updates)
- Some regedit changes
- Clean install with barebones apps

I should add that the machine never restarts itself as other users with the same error report. It always freezes, screen stays on, sound skips. I once waited 5 minutes to see if it would resume but no dice.

I have a replacement power supply on the way, but before I open that and try it, I figure I can get your thoughts. Maybe it's not the power supply at all. I do suspect it's a hardware issue but have been unable to confirm.

Thanks for reading!

Hello and welcome to TenForums :)

If you suspect hardware there are many useful tools you can try to stress test this aspect. I'll provide a series of links and guidelines here. With memory testing the most common mistake people make is that they do not test for long enough, 32GB of memory can take over a day to test to the full 8 or more passes.

Diagnostic Tests

 RAM TEST

Run MemTest86+ to analyse your RAM. MemTest86+ - Test RAM - Windows 10 Forums

   Note

MemTest86+ needs to be run for at least 8 complete passes for conclusive results. Set it running before you go to bed and leave it overnight. We're looking for zero errors here. Even a single error will indicate RAM failure.

 GPU TEST

Run Furmark to test your GPU. Furmark tutorial

   Note

Run Furmark for around 30 minutes

   Warning

Your GPU temperatures will rise quickly while Furmark is running. Keep a keen eye on them and abort the test if temperatures rise too high

 CPU TEST

Run Prime95 to stress test your CPU. Prime95 - Stress Test Your CPU - Windows 10 Forums

   Warning

Your CPU temperatures will rise quickly while under this stress test. Keep a keen eye on them and abort the test if overheating occurs.

 HDD TEST

Run HDTune to

• check the health,
• scan for errors, no quick scan but full scan
• run a benchmark.

It may take some time, but please take the time you need to perform it properly.
When above is done please make screenshots of the following

• the health,
• the error scan,
• the benchmark incl. following
  
  • transfer rate,
  • access time,
  • burst rate,
  • cpu usage.

Will run these overnight. Thanks.

OK. Ran pretty much everything:

No issues:

- Memtest: 8 Passes, no errors
- Furmark: 30 minutes, no issues
- HDTune: Health ok, no errors, good access and write speeds

Problems:

- Prime95: CPU temps go beyond 70C within a minute. End up having to shut it down. I'm running a stock heat sink. Will correct this soon.
- Used a multi meter and found the power supply to have some issues. 3.3V was a little low and the -12 volt pin was outputting -10. I've replaced the power supply.

Once I found the power supply issues I thought I had this in the bag. Alas, the system crashed the same way after 3 hours. But three hours is a lot longer than it was lasting before, so I think that was one issue. Next up I'll get a better heatsink and re-run the cpu test.

Managed to solve this.

Seems like I had a couple issues. The power supply was on the fritz and C-States needed to be turned off in the bios.

I went through all the steps listed on this thread by @mileskb
Windows 10 freezes randomly

I did change some of the power management settings in windows, but I had tried that before without results. It might have been a combination of C-States and power management.