• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Hardware issue - second opinions please

G

Geryon

n00b
Joined
Sep 23, 2009
Messages
48
Spec in sig. No OC on anything.

Symptoms:
  1. While gaming, the screen will turn black (but will not go on standby) and the PC will completely lock up, requiring a hard reset.
  2. After symptom 1 occurs, the PC will quite often do exactly the same thing again on the Windows login screen, or immediately after logging in. A second reset usually gets me back up and running.
  3. Occasionally, the game I've been playing (Dark Souls 3) will stop responding during map transitions and must be killed in Task Manager.
  4. After a couple of hours of general use (email, web browsing, listening to music, media playing, etc), random processes will start crashing all at once, quite often including dwm.exe (Desktop Window Manager). The system will then remain unstable until it is rebooted, at which point it resumes normal operation for another couple of hours.
Things I've tried:
  1. Clean Windows 10 installation on a spare SSD, with nothing else but drivers and Steam installed. Symptoms 1 and 2 still occurred. Couldn't really test symptom 4 since no other applications were installed.
  2. Various memory tests found no errors, although I've yet to do a full 8-pass Memtest86+ run.
  3. Reseated all the hardware components. Made no difference.
  4. Various stress tests to see what triggers symptom 1:
    1. OCCT with large and small data sets. No issues found.
    2. OCCT's Power test with AMD GPU. Triggered symptom 1 relatively quickly.
    3. OCCT's Power test with Intel integrated graphics. Triggered symptom 1 eventually, but took a lot longer.
    4. Furmark with various power limits set in the AMD driver. At +50% power limit it triggers symptom 1 within a few minutes. At default or negative power limits it still triggers symptom 1 but takes longer.
My initial feeling is that it could be my 13 year old Corsair HX620 PSU finally starting to fail, particularly since symptom 1 seems to be directly affected by the power limit set on the GPU, and I haven't seen any artefacting or BSODs that might suggest GPU or RAM issues.

However, symptom 4 doesn't seem like something that would be power-related (why would random processes suddenly start crashing after extended use while the OS remains responsive?), so I wonder if these problems could also be motherboard-related, or something else entirely.

Unfortunately, I don't have any spare parts available to swap out, so I'd have to start buying new ones to try.

Any advice before I start wasting money would be much appreciated!
 
I got about halfway through your post when the PSU came to mind. Fluctuating power can cause all kinds off oddness. And 13 years is certainly getting old for a PSU.

Another possibility is overheating. Have you checked the temps of the CPU/GPU? Heatsinks clean and the fans all working? When was the last time you reapplied thermal paste?
 
BlueLineSwinger said:
Have you checked the temps of the CPU/GPU?
Click to expand...

Yeah, sorry, I completely forgot to mention it but the temps are all fine. (Fans are running well, heatsinks aren't clogged, paste hasn't been changed for around 7 years but was top quality at the time).

I mean, they've always run a little hot because it's a warm room in general, but the CPU and GPU temps are nowhere near their thermal limits when these crashes occur.
 
up ram and cpu voltage a bit see if it hleps. if not the psu is your next step.
 
13 yo psu is where I would start. We did several 10 year re-tests on psu's and those are not something I would want in my system.
 
Well, a friend of mine mailed me a spare PSU to try. Same problem still occurred.

pendragon1 said:
up ram and cpu voltage a bit see if it hleps.
Click to expand...

Gave that a try. Made no difference.

Don't know what else I can do now. :(

I mean, if I have no way of proving where the fault lies, that leaves me with replacing random components until it works, or cutting my losses and doing a brand new build.
 
Geryon said:
that leaves me with replacing random components until it works, or cutting my losses and doing a brand new build.
Click to expand...
welcome to troubleshooting, can be hard without spare parts. but, break it down to the basic; board, chip, ram, cpu and psu. test all that. then add in the gpu and start testing that. see if you get anywhere.
 
Well I remembered I actually had a couple of old GPUs in my loft (a 7 year old HD 7950 and a 10 year old HD 5870!) so I've put the 7950 in my PC and the problems have all disappeared.

Dark Souls 3 is running fine with no crashes (albeit at a lower framerate), and no other processes are crashing, regardless of how long the PC is left running.

Seems like this was all a GPU issue after all, unless there was some weird interaction going on with my R9 390 and some other component.

Kinda surprised to be honest. First time I've ever seen a component just malfunction like that for no apparent reason. At least now I can take my time deciding what to do about it though.
 
Geryon said:
Kinda surprised to be honest. First time I've ever seen a component just malfunction like that for no apparent reason. At least now I can take my time deciding what to do about it though.
Click to expand...
Might have to do with that 13 YO PSU....
 
Oh, quite likely, I'm sure.

I think what I'll do is get a nice new Seasonic for now, and then do a complete new build later this year when the next-gen GPUs arrive.

Thanks for everyone's comments!
 
You must log in or register to reply here.
Back
Top