[SOLVED] Did my PSU fail and fry my GPU in the process?

Sep 10, 2021
3
0
10
I'm wondering if I have one problem or two.

Yesterday I noticed my computer would spontaneously turn off and then restart. I think it happened about 3 times in total after opening both YouTube and a game. Once it happened right as I tried to boot up Sibelius. Since I cleaned my graphics card and CPU two weeks ago and both were running at normal temperatures (CPU between 40C and 60C, GPU no higher 80C if a game was open), I suspected my 10-year-old PSU was the culprit. Since both my CPU and GPU failed about two weeks ago from overheating, I kept an eye on metrics and tried to induce failures, but both were functioning normally up until the computer shut itself off and rebooted. However after those 3 failures I had something else happen on the fourth boot - the BIOS screen would appear as normal but the Windows logo would flicker at the end, and the login screen would look like it was in 256-color mode. Doing nothing would make the entire screen wipe to a shade of blue, and logging in would cause all kinds of crazy colors and at least one of my monitors to fail. The whole system would become unresponsive within 60 seconds no matter what I did.

Going back to the first issue I had, I subbed in the new PSU but I'm still having the issue with the colors. Any sort of boot with the default graphics (safe mode, delete all drivers from my video card) will work as normal, but if I try to install any set of drivers in the past 12 months (tested 4 different ones) I will get the colors issue. Did I somehow fry the card, or did I misdiagnose the issue altogether?

Old PSU: Corsair GS600
New PSU: Corsair RM650x
GPU: Zotac NVIDIA GTX970
CPU: Intel i7-4790K
Motherboard: Gigabyte GA-H97M-D3H

Aside from the PSU's, all of these parts were purchased in 2015 and have worked without any major incident up until this past month.
 
Solution
I'm wondering if I have one problem or two.

Yesterday I noticed my computer would spontaneously turn off and then restart. I think it happened about 3 times in total after opening both YouTube and a game. Once it happened right as I tried to boot up Sibelius. Since I cleaned my graphics card and CPU two weeks ago and both were running at normal temperatures (CPU between 40C and 60C, GPU no higher 80C if a game was open), I suspected my 10-year-old PSU was the culprit. Since both my CPU and GPU failed about two weeks ago from overheating, I kept an eye on metrics and tried to induce failures, but both were functioning normally up until the computer shut itself off and rebooted. However after those 3 failures I had something else...
I'm wondering if I have one problem or two.

Yesterday I noticed my computer would spontaneously turn off and then restart. I think it happened about 3 times in total after opening both YouTube and a game. Once it happened right as I tried to boot up Sibelius. Since I cleaned my graphics card and CPU two weeks ago and both were running at normal temperatures (CPU between 40C and 60C, GPU no higher 80C if a game was open), I suspected my 10-year-old PSU was the culprit. Since both my CPU and GPU failed about two weeks ago from overheating, I kept an eye on metrics and tried to induce failures, but both were functioning normally up until the computer shut itself off and rebooted. However after those 3 failures I had something else happen on the fourth boot - the BIOS screen would appear as normal but the Windows logo would flicker at the end, and the login screen would look like it was in 256-color mode. Doing nothing would make the entire screen wipe to a shade of blue, and logging in would cause all kinds of crazy colors and at least one of my monitors to fail. The whole system would become unresponsive within 60 seconds no matter what I did.

Going back to the first issue I had, I subbed in the new PSU but I'm still having the issue with the colors. Any sort of boot with the default graphics (safe mode, delete all drivers from my video card) will work as normal, but if I try to install any set of drivers in the past 12 months (tested 4 different ones) I will get the colors issue. Did I somehow fry the card, or did I misdiagnose the issue altogether?

Old PSU: Corsair GS600
New PSU: Corsair RM650x
GPU: Zotac NVIDIA GTX970
CPU: Intel i7-4790K
Motherboard: Gigabyte GA-H97M-D3H

Aside from the PSU's, all of these parts were purchased in 2015 and have worked without any major incident up until this past month.
I'd remove the GPU, move the monitor cable to the motherboard connection and boot with the Intel integrated graphics. If it works fine then the GPU would be highly suspect. Last time I had color band issues it turned out to be the monitor cable itself so you could try swapping it out first. Even if the PSU was not the root cause of your problem, it was time to replace it so don't consider it wasted money.
 
Solution
Sep 10, 2021
3
0
10
I don't think it's the monitor cables. I've got 3 screens (Dell U2415) and the cables all have different connector types. I'm seeing the color issue on all 3 when I try to load any driver.
 

rakibfahadgts

Reputable
Aug 4, 2018
145
14
4,615
A PSU more likely wouldn't degrade color output. It seems like the GPU. Because if the GPU driver crashes PC would restart.

On a different note why are you using a 10 year old PSU. It's the first thing that you want to change within 5 years. Rest are dependant on your preference
 
Sep 10, 2021
3
0
10
Re: 10 year old PSU: I mean it worked this far, so I don't see what would have been the point in replacing it early.

Given that DDU + fresh driver install doesn't seem to fix the problem, is getting a new GPU my only way forward?
 

Vic 40

Titan
Ambassador
You cleaned the gpu, might want to recheck if all is well there, maybe give all parts on the pcb a good rub with iospropyl alcohol with out destroying something of course and if there are thermal pads replace those too.

Can also try even older drivers.

And with a gpu replacement if necessary would i suggest to get a better/new psu too.