Sudden PC crash, wouldn't boot and fans didn't spin for 10 minutes

techdeck12

Reputable
Jan 29, 2015
19
0
4,510
Hi guys, I recently had an issue while I was playing a video game. The PC crashed after a few hours of playing (suddenly turned off as if the power cut out). When I tried turning it back on, the HDD powered on, but the fans wouldn't spin and PC wouldn't boot.

When I opened the side panel and tried again, debug light indicators on the motherboard blamed the CPU for the issue. 10 minutes after the initial crash as I was putting the RAM in different DIMM slots, I pressed the power button again, the fans started spinning and I was able to boot to BIOS.

Currently everything seems fine. I'm typing this post on the PC that crashed. My question is how else can I troubleshoot this issue? I ran Passmark's Memtest86 and it showed 0 errors. My guess is that the CPU overheated and caused itself to turn off. What do you think happened?

My rig:
MSI z170a-pro
CPU Skylake 6600k overclocked to 4.6ghz (tested stable with Prime95 and other stress tests)
XFX Rx 480 GTR edition video card
1x8GB RAM stick 2400Mhz
Corsair RMx 850 Watt gold PSU
 

spat55

Distinguished
Check the temperature of the CPU when under load. That does sound like a CPU overheat as the motherboard is saying it was caused by CPU. I wouldn't of thought it would be the PSU but you can never really rule that out.
 

techdeck12

Reputable
Jan 29, 2015
19
0
4,510


OK I can try running Unigine Heaven stress test for a while and check the temps. I started the game that the PC crashed on, and the temps were around 50 degrees C only after 5 minutes.
 

spat55

Distinguished


Could be an unstable overclock maybe, have you tried running LinX or IBT for a few hours? That is my go to for checking overclocks. Just watch the temperature for the first 5 minutes though as it can and will get hot fast.
 

techdeck12

Reputable
Jan 29, 2015
19
0
4,510


I haven't tried IBT because it scares me. However I tested the overclock on P95, x264, and the CPU-ID stress test
 

spat55

Distinguished


What is wrong with IBT? Just make sure the temps don't go too nuts. How long have you ran P95? Thing is I've done P95 for 6 hours before then ran IBT for 10 minutes and got a crash, so I don't use it unless finalizing a overclock.

IIRC don't go over 85c, hwmonitor will go red when near TJmax
 

techdeck12

Reputable
Jan 29, 2015
19
0
4,510


I keep hearing that it's way too stressful for the CPU. However I'm getting over my fear and currently running it at increasing stress levels. I remember running P95 for up to 10 hours without a crash, so I assumed that my overclock is stable.
 

techdeck12

Reputable
Jan 29, 2015
19
0
4,510


Ok I ran Intel Burn Test for 20 loops at Very High setting, the cpu passed with no issue, which means my overclock should be stable. It's there something else I can try? The crash seems to be a random event so far.
 

spat55

Distinguished


How old is the PSU? That could be something to think about. I'd consider running OCCT and run the PSU side of it, it runs the GPU/RAM/CPU and everything to the max to see if it will crash.
 

techdeck12

Reputable
Jan 29, 2015
19
0
4,510


I got my PSU in the past several months from New Egg, so very new. I'll try OCCT out and see how it goes. Thanks for helping me out so far, it's very much appreciated.
 

spat55

Distinguished


No problem it's a horrid feeling having something you spend good money on not working correctly.
 

techdeck12

Reputable
Jan 29, 2015
19
0
4,510


I ran OCCT for two separate 20 minute intervals. The first run, OCCT ran into an error and stopped testing. I ran a second time, but stopped it manually after 20 minutes because the CPU temps were getting too high for my liking (peaked at 83 degrees for Core0 and Core1). I'm currently running the Intel Diagnostic Tool.

Edit: Just finished the diagnostic. Everything passes. My PC crash seems like a one-off occurrence so far.
 

techdeck12

Reputable
Jan 29, 2015
19
0
4,510
I ran OCCT for a 3rd time and the PC completely shut down. Currently if I try turning it on, the fans don't spin and the debug light on the mobo points to the CPU. It looks like I was correct in guessing that the CPU overheated and shut itself off. There's nothing I can do now besides to wait for the CPU to cool down so I can boot it up.

It looks like the only thing I can do is monitor the CPU temp while I'm playing a video game and make sure the temps don't reach 80 Degrees Celsius. If they do, then I'll have to lower the clock speed and the vcore voltage.
 

spat55

Distinguished


80c isn't that hot for skylake tbh, awhile on the high side I'd say it could be the PSU. Next time you log on I know it won't be that accurate but use HWmonitor and look at the +12v and +3.3v along with +5v and see what it says. Leave it running awhile gaming and see if they drop to low levels. I have a weird feeling it could be PSU.

The CPU would cool down to safe levels within a second without load so I don't believe it's that.
 

techdeck12

Reputable
Jan 29, 2015
19
0
4,510
I tried a tip that I found elsewhere on this forum. It was a bad CMOS battery! I replaced it with a brand new one from the store, and now it's booting into the OS. Thanks to spat55 and this website for the help!