Jump to content
Corsair Community

How can I tell if my PSU killed my GPU or not?


K1llrzzZ
 Share

Recommended Posts

Hey!

 

So about two weeks ago my 2080 Ti died, and I started the RMA process. It's "death" was kind of weird: A few months before it gave up it was already acting up, I turned on my computer, it booted up fine, but I got no signal to my monitor. I tested everything, turns out, my GPU was not recognised but when I connected the display port to my mobo, it gave a signal, I was in Windows and it was like I had no GPU other then the integrated one. A couple of reboots solved the problem, and once the problem was solved, my PC actually worked fine. Once my GPU was recognised, I could play for hours, no crashes to desktop, no BSODs, no random PC reboots, no artifacting, nothing out of the ordinary.

 

This happened a couple of times, but a few reboots always fixed the issue. Then it stopped happening for like 1-2 months before it happened again. It happened, 2 reboots and it's fixed. Then the next day I tried it 5 times... nothing.

 

So I took my computer to a friend of mine who works at a computer store, funny enough, my gpu actually worked at first, but then he updated my mobo BIOS and it finally started beeping signalling that the GPU is dead.

 

We tried putting another GPU in the PCIe slot to test out if it's working, and it did.

 

I'm using my PC with the integrated GPU since, no issues, my PSU is working, my motherboard is working. My question is: Was my GPU faulty or did my PSU killed it? Or my mobo? How could I tell? Because if I get a replacement or a new card I don't want it to die again, and if it will I'm pretty sure I'll lose the warranty too, and they will say it was my fault for using it with a faulty PSU.

 

My GPU was an Asus RTX 2080 Ti DUAL OC I bought back in november 2018, it has the Micron VRAM on it, and I know there were a lot of faulty 2080 Tis back then, but they all died pretty soon and they had this weird artifacting before they did which looked like space invaders. Mine worked fine for about 1.5 years, it still worked fine IF my PC decided to recognize it, no "space invaders".

 

My PSU is a Corsair RM850 I bought back in late 2014, so it's almost 6 years old now, but I had no issues with it so far. Also like I said, everything else still works fine in my PC. I had an SSD die last summer, but it might have been because of the heat, it's really hot in my room and it was between my GPU and my CPU so it got really hot.

 

My mobo is an Asus Strix Z390-H Gaming, and my CPU is an i9 9900K. Is it possible the the motherboard is faulty and that killed the GPU?

 

Thanks for the help and sorry for the long post.

Link to comment
Share on other sites

PSU's kill other parts when they fail on the secondary side. You're saying your PSU did not fail.

No it's still working. I'm no expert, I was just thinking, maybe as it's aging it doesn't regulate voltages that well anymore, and gave my GPU a little bit too much one time which damaged it. I don't know if this is possible.

Link to comment
Share on other sites

  • Corsair Employees
No it's still working. I'm no expert, I was just thinking, maybe as it's aging it doesn't regulate voltages that well anymore, and gave my GPU a little bit too much one time which damaged it. I don't know if this is possible.

 

That's actually a VERY valid argument and is often times quite true.

 

But your PSU isn't really that old and I would expect to see other related damage.

 

You could shine a light into the fan grill directed at where the smaller caps are (on the DC output side opposite of where the big AC caps are) and see if any of them are "swelling" (the top will "dome" a bit).

Link to comment
Share on other sites

That's actually a VERY valid argument and is often times quite true.

 

But your PSU isn't really that old and I would expect to see other related damage.

 

You could shine a light into the fan grill directed at where the smaller caps are (on the DC output side opposite of where the big AC caps are) and see if any of them are "swelling" (the top will "dome" a bit).

 

A took a couple of pictures, you can see inside the PSU from the back where the power connector is, and from the bottom where the fan is.

 

I have an old CoolerMaster HAF932 case, and the fan can blow the air out from the bottom, and it has wheels so there's a good 5 cm (~2 inches) between the floor and the case. I'm just telling this in case it's not regular for the fan to be looking down, I didn't build this PC.

 

Here are the pictures:

 

https://imgur.com/a/HCNpzpN

 

I hope they are at least somewhat useful, it's hard to make good pictures, even with a naked I you can barely see in, especially at the bottom where the fan is, it covers up most of the stuff inside and I'm definitely not qualified to take it apart.

 

Can you tell something from these pictures? Does it look ok?

 

I also read that you can test a PSU with a digital multimeter, the person who built my PC works at a PC hardware shop, I'm sure they have one, could that determine if there's something wrong the the voltage regulation of the PSU or not?

Link to comment
Share on other sites

  • Corsair Employees

I'm afraid I can't see anything in those pictures.

 

Also, a multi-meter can't really be used to measure ripple. At least not your typical DMM.

 

If you're concerned, and all of your parts are quite new EXCEPT FOR the PSU, then why not just replace the PSU? Seems like you've replaced everything else.

Link to comment
Share on other sites

Well my mobo and my CPU are 2 years old, my SSD is even newer because I had it RMAd, so that's also something that makes me worried, but it could have been because of the heat. I'm planning to keep those for another 1-2 years, I game at 4K so they should be enough. My case is like 10 years old, my 2TB HDD is also 5-6 years old so not everything is now.

 

If I know there's something wrong with my PSU then I'll obviously replace it, I don't want to risk it damaging anything else, but if it's fine then I wouldn't want to waste my money on a new PSU for no reason, that's why I'm asking around. I guess I'll ask my friend to test it before getting the new graphics card, hopefully he can tell if it's in good shape or not.

 

Good quality PSUs have protections that shouldn't let voltage damage other components even as they're aging right? I read that this is mostly an issue with cheap brands. I don't know if these RM850 PSUs from 2014 are known for having issues like that.

Link to comment
Share on other sites

  • Corsair Employees

Good quality PSUs have protections that shouldn't let voltage damage other components even as they're aging right? I read that this is mostly an issue with cheap brands. I don't know if these RM850 PSUs from 2014 are known for having issues like that.

 

Voltages, yes.

 

But I'm not talking about voltages. The secondary capacitors filter ripple. There's no protections for that.

 

Since the PSU is out of warranty, maybe your friend (if he knows what he's doing) can open it to do a visual inspection.

Link to comment
Share on other sites

Well he's a PC repair guy so he knows what he's doing, definitely better then I do. Also I googled PSU ripple and apparently it could also be tested with an oscilloscope, but I have no idea if they have one of those, hopefully they do.

 

Anyways, thank you for your help!

Edited by K1llrzzZ
Link to comment
Share on other sites

They just got back to me about my GPU, they tested it, turns out, it works fine. So the good news is, that my PSU didn't kill my GPU. The bad news is, that it might be faulty, and possibly the reason why my GPU wasn't recognised. Hopefully getting a new one will fix the issue... unless something else is the problem. I really hope I won't have to replace my entire system before I could get it working properly again.
Link to comment
Share on other sites

My PSU, since my GPU is fine apparently. I got a good deal on a PSU, it was 30% cheaper because the box was opened, brand new, same warranty and everything, all cables there, so I ordered it. It's a Corsair HX1200. Maybe a bit overkill but I expect it to last a few builds and 1200W will be more then enough for the next-gen GPUs.
Link to comment
Share on other sites

My PSU, since my GPU is fine apparently. I got a good deal on a PSU, it was 30% cheaper because the box was opened, brand new, same warranty and everything, all cables there, so I ordered it. It's a Corsair HX1200. Maybe a bit overkill but I expect it to last a few builds and 1200W will be more then enough for the next-gen GPUs.

 

You will be well served with the HX1200 which has capacity for even the RTX 3090 etc

Link to comment
Share on other sites

It may be just me but I often see the PSU being blamed for stuff when in my experience PSU's are highly unlikely to be the cause of problems.

 

Components rarely commit murder. They more often just commit suicide. Sometimes when they do that, they can take out other things but even that is unusual.

 

It sounds to me like you have the dreaded warm-up issue. This is when a PC is cold, it doesn't work, but once it has warmed up a little, it works just fine. This unfortunately just gets worse and worse until one day it doesn't power up at all.

 

The offending component could be anywhere but it is most likely on your motherboard, since your graphics card has been tested already. Unfortunately these problems are really hard to track down.

 

Someone once told me that warm-up death is due to a failing capacitor somewhere in the chain, but I am not an expert at identifying individual components

Link to comment
Share on other sites

 Share

×
×
  • Create New...