[SOLVED] PC restarts during gameplay. No issues when running stress tests.

[dolokhov]

Problem:
PC started to suddenly freeze up during use - then would restart on it's own. Usually happens within first 5 mins, almost always when a game is loaded to full screen.


SPECS:

 

[dolokhov]

UPDATE - Looks like the PC still restarted, but it lasted much longer than usual and was able to run games.

I am going to try re-seating my memory. Also going to try running it without the dog integrated.

Any other suggestions are appreciated. I would hate to RMA the PSU or GPU without ruling out other causes.

 

[Grobe]

Can you perform a stress test using OCCT and share screen shots ?

 

[dolokhov]

Can you perform a stress test using OCCT and share screen shots ?

Grobe, thanks for your time.

Here are some screenshots:

 

[dolokhov]

Power test:

 

[dolokhov]

I wonder if this is a symptom of an issue - I started Magic the Gathering Arena app, and I noticed my fan rpms seem to be spiking up then turning off:

Experienced a shutdown about 5 minutes after this screenshot. Temps were fine.

When it crashes, I do hear a sound - I am thinking it's coil whine. I hear it when the PC is restarting and immediately after, then it goes away.

 

[Grobe]

post #4 - could you redo and select "+12V" (third image) this time ?

 

[dolokhov]

post #4 - could you redo and select "+12V" (third image) this time ?

I think this is what you are looking for Grobe:

Does the coil whine I hear during and after restart provide any clues? It's hard to say which part it's coming from, I could lay the all out the table and find out. I only hear that whine at times when it crashes.

 

[dolokhov]

Fan once MTG Arena loaded:

Crashed about 10 mins into game.

 

[Grobe]

Coil whine doesn't prove anything - other that there is a coil somewhere that when under certain condition when there is a mechanical resonance that make the winding vibrates within audible frequencies. Or it may be due to bad design of the pcb if under some circumstances, the coil actually overloads.
Also - you probably wouldn't be able to locate the exact position of the coil in question.

Ok - back to your images. It indicates PSU are ok. The images of fan speeds are of no interest because it all depend on the setup/configuration of the fan controller (doesn't matter if hw or software).

 

[dolokhov]

Coil whine doesn't prove anything - other that there is a coil somewhere that when under certain condition when there is a mechanical resonance that make the winding vibrates within audible frequencies. Or it may be due to bad design of the pcb if under some circumstances, the coil actually overloads.
Also - you probably wouldn't be able to locate the exact position of the coil in question.

Ok - back to your images. It indicates PSU are ok. The images of fan speeds are of no interest because it all depend on the setup/configuration of the fan controller (doesn't matter if hw or software).

Any theories on what is causing the restarts Grobe?

 

[dolokhov]

I've decided on a plan. See below and let me know if I am off base.

1. Test RAM - Tested via Windows tool, no issues.
2. Thorough clean of fans, PSU and replace thermal paste. I noticed the heat sink near the CPU moves a little, and the back of the board where the heatsink is gets quite hot. Reason for concern?
3. New PSU. I am going to buy a cheap PSU that can serve as a backup in the case where switching it does not stop the problem. If it is the PSU, then I will RMA my EVGA G2 750.
4. New GPU. I'll pick this up off Amazon so I can potentially return it.

 

[Grobe]

Actually, I'm more into failure of motherboard or GPU.

Try to uninstall current GPU driver, and install an older version - just to see if that fix the issue.

Also, there are some people suggesting the software Catzilla does a more realistic stress test of GPU. You can try that too (both with current and older driver) - to see if it runs successfully.

 

[dolokhov]

Actually, I'm more into failure of motherboard or GPU.

Try to uninstall current GPU driver, and install an older version - just to see if that fix the issue.

Also, there are some people suggesting the software Catzilla does a more realistic stress test of GPU. You can try that too (both with current and older driver) - to see if it runs successfully.

Thanks for your help Grobe.

For anyone who might find this thread while googling - I never did find the reason. I am planning on rebuilding my aging system or getting a laptop.

 

[dolokhov]

Update in case anyone reads this thread at some point with similar problems:

I ended up selling the system, and the person who purchased it said they were able to repaste the VRAMS (?) on the GPU and reseat the CPU and they can now run it with stability.

 

[Grobe]

Thanks a lot for sharing 😀