Crash to auto restart, CPU overheat? PSU? GPU? How to check?

panadtd

Commendable
Sep 29, 2016
3
0
1,510
I've started to get crashes that makes my PC go black screen (no BSOD) then restart during games about 4 months ago, it was only in occasionally in Dying Light. But then it happened pretty often in Overwatch, and lately Attack on Titan, and now FIFA17.

Here is part of my spec:
MB: MSI Z97-G55 SLI
CPU: Intel i7-4790K
GPU: MSI GTX970 Gaming 4GB
RAM: Kingston HyperX DDR3-1866 8GBx2
SSD: Crucial M550 256GB
HDD: WD Black 640GB/ WD Red 3TB
PSU: Highpower EP-750S
No OC at all.

I've read many threads which points the problem to CPU overheat in general, and indeed I've started monitor it's temperature. it sits at 40-45C idle, and above 70 on load, maxing at 83 (CPUID Monitor). I've already reinstalled the stock heat sink (and reapplyed paste) 3 times, tried using normal amount of paste, extra amount, and less ('coz it ran out, haha). The temp. is more or less the same.

I'm not sure what is the temp. exactly when it crashes because I'm usually playing in fullscreen when it crashed. But I think 90% of the time it crashed some kind of a transition, like just when a game is just won in Overwatch, or just when the games go from menu to cutscene, from cutscene back to normal gamplay, etc. Which led me to wonder if it as anything to do other parts like PSU / GPU.

I had GPU overheating with my last rig, and it usually started to get video lag and maybe audio loop before it crashes/ hangs. And my GPU is sitting at 71C with FIFA running (menu) in the background. And the fact that it never crashed in some other high requirement games like Far Cry 4 made met feel it's not the GPU (could be wrong).

I'm suspicious about the PSU, it's probably not a quality one. I'm not sure if it's unstable power making the CPU heat up, or if it's simply the PSU that's causing the crashes.

Or maybe it can even be the other parts? So I'm mainly asking, what so I do/check to pin point the problem? From your experience, what does it look like that's the cause?

I do plan to buy a new PSU to try as the first step, I'm also thinking about getting a better CPU cooler. But I just fear that if it's not the problem then it'll be a waste.

Thanks in advance.
 
Solution
Crashes during transition do not point to PSU problems or overheating, because both of those would trigger the problem at the most intense moments (lots of effects, lots of things to render et cetera) and also happen in other games.

Crash during game transitions could point to RAM problems or storage drive problems (because it could either be the RAM being loaded with new data or the drive accessing that data). The fact that it only happens in some games points more to the storage drive because a RAM problem would also affect other tasks and would also cause this problem (albeit rarely) even on the Windows desktop.

Run Memtest just to be sure and also check your drive's health with CrystalDiskInfo.
Crashes during transition do not point to PSU problems or overheating, because both of those would trigger the problem at the most intense moments (lots of effects, lots of things to render et cetera) and also happen in other games.

Crash during game transitions could point to RAM problems or storage drive problems (because it could either be the RAM being loaded with new data or the drive accessing that data). The fact that it only happens in some games points more to the storage drive because a RAM problem would also affect other tasks and would also cause this problem (albeit rarely) even on the Windows desktop.

Run Memtest just to be sure and also check your drive's health with CrystalDiskInfo.
 
Solution

panadtd

Commendable
Sep 29, 2016
3
0
1,510


I got the Cooler Master V750 in the end. I believe 650 would be enough, but they were out of stock =/

For the cooler, I wanted to get the Cooler Master Hyper 212 EVO, but again, it's sold out and I could only get a Hyper LED. The CPU still idles at around 40C, but I'm no longer seeing 80 at max. For the most part it is around 65-73 on load. Considered that room temp is generally quite warm here, I think I will accept these temp.



HDDs are at 95% healthy according to CrystalDiskInfo. And I didn't have time to get Memtest to work, but I used the mdsched.exe command and the rams seemed to be fine.

Anyway, I got a new PSU, and after installing it, my PC failed to boot. I got it to boot again after removing 1 of the 2 rams. Then I tried to boot with only the other ram and it worked too. So I suppose both rams are actually still good, but maybe the ram slot 3 (I used to put rams in 1/3 for dual channel) is the problem. As for now I'm putting them in1/2.

I'm just wondering if slot 3 is indeed the problem, how did the PC manage to boot and run for so long before. I may test with the rams more later, but for now I've been playing the games mentioned before without any crash for 2 days.

Thanks for the help again.

 
Since the RAM is running fine in slots 1/2 but not 1/3 then it means either of two things:

1)Slot 3 is broken.

2)CPU's IMC is subpar (somehow slipped through QC) and can't handle dual-channel at stock voltage.

To verify which it is, place them in slots 2/4. If they work in 2/4 then Slot 3 is the culprit.
 

panadtd

Commendable
Sep 29, 2016
3
0
1,510


Thanks, will do the next time i open the case up. But slot 4 is actually slightly blocked by the fan on the cooler, I guess I could put the fan on the other side of the cooler to try.