Random PC crash since 4k monitor (no BSOD)

Jun 18, 2018
5
0
10
Recently I have bought a 4K monitor and replaced one of my two 1080p monitors with it. That same evening, my PC randomly shut down and I was unable to restart it. I tested the PSU (which was cheap) and found it was dead, so I replaced it with a better one.

Soon after, I started getting the random crashes again (average once or twice a week). I checked the event log, but saw nothing critical except for event 41, which is rather obvious for this case.

I also checked the wall outlet with a power meter and my max draw was around 300W. I hooked up the PSU via USB to Corsair Link and saw that max draw was around 270W (my PSU is 650W). The 3.3v, 5v and 12v are within spec and seem fine. I asked Corsair Link to keep a log of the metrics to see if I can see anything weird at the time of crash, but all values seem normal at the time of crash.

Temperatures are fine (CPU 60-70c, GPU 70c while gaming including VR).

Here are some metrics from Corsair Link: https://docs.google.com/spreadsheets/d/e/2PACX-1vRetO3iQY3gbCfvCWI9G3JE_sXcyVIHErnwkYEgQcn4Rx1G07vcE80VlEUobYIgarV7fWGUh3bkDvtL/pubhtml

I cannot reproduce the issue on demand; I have to wait until it happens. I've stress tested using Prime95, Furmark and AIDA64. I have no run memtest, but Prime95 blend was fine.

CPU: i7-7700
Motherboard: Asus B250M-A
GPU: NVidia GTX 1080
PSU: Corsair RM650i (650W)
RAM: 2x8GB DDR4-2133

I would appreciate any help diagnosing the issue. Thanks!

Edit: I have also upgraded from Win7 to Win10 since this issue has started happening. I did a clean install on another SSD. I installed newer video drivers from NVidia with no luck.
 
Jun 18, 2018
5
0
10
Well, it's the same. The PC shuts off as if the power goes out. I have considered the possibility that some component is damaged, I'm trying to find which component that is.

I can give memtest a run, although I ran the memory stability test through Aida64 and I believe Prime95 blend covers some RAM too.

I'll let you know the results.
 
Jun 18, 2018
5
0
10
I've used both RealTemp and Corsair Link. I can maybe find another modular PSU that I can try, not sure. I don't have a spare one that I can try right away. Even if I do, I can't be 100% sure since it's very random and not reproducible (at least as far as I can tell).
 
I generally use RealTemp myself for Intel CPUs. Since you tried clean OS installs on multiple drives, it eliminates that. Although should be unrelated, have you tried not using the 4k monitor, or using a different display output for the 1080? Do you have any other spare parts laying around to narrow this down. Right now, I'm still leaning on either a faulty replacement PSU, or damaged component from the failed cheap PSU.
 
Jun 18, 2018
5
0
10


Memtest reports no errors. I tried running single monitor (both 4k and 1080p) and the crash still occurs but not necessarily while gaming (I've had a couple crashes where the PC was completely idle).