[SOLVED] PSU stress test fried something?

Aug 20, 2021
4
0
10
Over the past few days while I've been playing games, every once in a while the pc would either crash to desktop or black screen and restart itself. Well today it was black screening within minutes so I decided to stress test the PSU, and it black screened within ten seconds and wouldn't turn back on. The fans and lights will flicker for a second when trying to turn it on then not let me try again until I've unplugged the PSU.

I've removed piece by piece to see what the issue was, and I've found that if the GPU is plugged in and ONLY has the six pin connector in the GPU, it will turn on, but will not if it has both the six pin and the eight pin plugged in. I've tried swapping the wires out as my PSU is modular but no luck. The eight pin is also connected though an adapter for two six pins if that has to do with anything.

Just to clarify, the GPU needs 1 six pin and 1 eight pin. My modular PSU is using an eight pin to six pin for the GPU's six pin slot. The other wire is another eight pin to 2 six pins, which are then attached to an adapter that takes 2 six pins and outputs the eight pin for the second GPU slot.

Right now I'm thinking that I somehow shorted either my PSU or my GPU, but I don't have any spare parts to test with so I cant tell. Both of these parts are pretty old so I wouldn't be surprised if one of them died. Currently I'm leaning toward the GPU being <Mod Edit> as games have been giving me a "heads up" that a game is about to crash by making one of my other monitors grey screen (then the game crashes within 30 seconds) for a while now. But my power usage in my room commonly trips the house's breaker switch and I have to go reset it in the basement, so I imagine it's taken a toll on my PSU somehow. Any idea where the problem could be?
The GPU is a nvidia 980 Ti and the PSU is a evea supernova 850 G2 gold power supply.

EDIT: So I found an old Geforce 550 Ti and I managed to get it started up at the very least. It only needed one six pin from the PSU if that matters. Any recommendations I should take at this point? Should I try the stress test again to see if it shuts down again? At this point I'm unsure if it would screw up this GPU as well.
 
Last edited by a moderator:
Solution
Honestly, I'd suspect the motherboard. It's bad enough that you're running a 125W CPU on a bargain AM3+ motherboard that's notorious for frequently having trouble handing those at stock, but you actually sought to stress it. I wouldn't be surprised if some VRMs gave up the ghost.

And what do you mean by "eight pin to six pin?" Are you talking about the pins on the side of the cable that connects to the modular power supply? Or some kind of adapter? If you're talking about the side that connects to the modular PSU, those are proprietary and the numbers don't matter and if you're talking about adapters, you should never be using adapters with PSUs. I'm asking because it's important to have all the facts straight at the start (this is...
Over the past few days while I've been playing games, every once in a while the pc would either crash to desktop or black screen and restart itself. Well today it was black screening within minutes so I decided to stress test the PSU, and it black screened within ten seconds and wouldn't turn back on. The fans and lights will flicker for a second when trying to turn it on then not let me try again until I've unplugged the PSU.

I've removed piece by piece to see what the issue was, and I've found that if the GPU is plugged in and ONLY has the six pin connector in the GPU, it will turn on, but will not if it has both the six pin and the eight pin plugged in. I've tried swapping the wires out as my PSU is modular but no luck. The eight pin is also connected though an adapter for two six pins if that has to do with anything.

Just to clarify, the GPU needs 1 six pin and 1 eight pin. My modular PSU is using an eight pin to six pin for the GPU's six pin slot. The other wire is another eight pin to 2 six pins, which are then attached to an adapter that takes 2 six pins and outputs the eight pin for the second GPU slot.

Right now I'm thinking that I somehow shorted either my PSU or my GPU, but I don't have any spare parts to test with so I cant tell. Both of these parts are pretty old so I wouldn't be surprised if one of them died. Currently I'm leaning toward the GPU being <Mod Edit> as games have been giving me a "heads up" that a game is about to crash by making one of my other monitors grey screen (then the game crashes within 30 seconds) for a while now. But my power usage in my room commonly trips the house's breaker switch and I have to go reset it in the basement, so I imagine it's taken a toll on my PSU somehow. Any idea where the problem could be?
The GPU is a nvidia 980 Ti and the PSU is a evea supernova 850 G2 gold power supply.

EDIT: So I found an old Geforce 550 Ti and I managed to get it started up at the very least. It only needed one six pin from the PSU if that matters. Any recommendations I should take at this point? Should I try the stress test again to see if it shuts down again? At this point I'm unsure if it would screw up this GPU as well.
Give us parts list including PSU details.
 
Aug 20, 2021
4
0
10
Motherboard: Gigabyte GA-78LMT-USB3 FA

Power Supply: EVGA SuperNOVA 850 G2 Power Supply
P/N: 220-G2-0850-XR
AC Imput: 100-240 VAC, 10A, 50-60 hx
DC Output: |+3.3V| +5V | +12V | +5Vsb | -12V |
Max Current: | 24A | 24A | 70.8A \ 849.6W| 3.0A | 0.5A |
| 120W | 849.6W | 15W | 6W |
Total: | 850W @ +50C |
1x 24 PIN ATX
2x 8pin (4+4) EPS (CPU)
4x 8pin (6+2), 2x 6pin PCIE
10x SATA
4x Four-pin Peripheral (Molex)
1x Floppy


Processor:
AMD FX(tm) - 8350 Eight Core Processor


Graphics Card:
Nvidia GeForce GTX 980 Ti (?4gb?)
http://www.geforce.com/hardware/desktop-gpus/geforce-gtx-780/specifications


Hard Drives:
Western Digital WD10 EALX-009BA0 SCSI Disk Device
OCZ Vertex 460 Solid State Drive


RAM:
HyperX 2x8GB Dual DDR3 1333
HyperX 1x8GB DDR3 1333


CD-ROM Drive:
ATAPI iHAS124 C SCSI cdRom Device


Case/Tower:
IBUYPOWER i-Series 301


OS:
Windows 10 Home Premium 64-bit
 
Over the past few days while I've been playing games, every once in a while the pc would either crash to desktop or black screen and restart itself. Well today it was black screening within minutes so I decided to stress test the PSU, and it black screened within ten seconds and wouldn't turn back on. The fans and lights will flicker for a second when trying to turn it on then not let me try again until I've unplugged the PSU.

I've removed piece by piece to see what the issue was, and I've found that if the GPU is plugged in and ONLY has the six pin connector in the GPU, it will turn on, but will not if it has both the six pin and the eight pin plugged in. I've tried swapping the wires out as my PSU is modular but no luck. The eight pin is also connected though an adapter for two six pins if that has to do with anything.

Just to clarify, the GPU needs 1 six pin and 1 eight pin. My modular PSU is using an eight pin to six pin for the GPU's six pin slot. The other wire is another eight pin to 2 six pins, which are then attached to an adapter that takes 2 six pins and outputs the eight pin for the second GPU slot.

Right now I'm thinking that I somehow shorted either my PSU or my GPU, but I don't have any spare parts to test with so I cant tell. Both of these parts are pretty old so I wouldn't be surprised if one of them died. Currently I'm leaning toward the GPU being <Mod Edit> as games have been giving me a "heads up" that a game is about to crash by making one of my other monitors grey screen (then the game crashes within 30 seconds) for a while now. But my power usage in my room commonly trips the house's breaker switch and I have to go reset it in the basement, so I imagine it's taken a toll on my PSU somehow. Any idea where the problem could be?
The GPU is a nvidia 980 Ti and the PSU is a evea supernova 850 G2 gold power supply.

EDIT: So I found an old Geforce 550 Ti and I managed to get it started up at the very least. It only needed one six pin from the PSU if that matters. Any recommendations I should take at this point? Should I try the stress test again to see if it shuts down again? At this point I'm unsure if it would screw up this GPU as well.
I would suspect the PSU you should try another known working PSU in the system.
 

DSzymborski

Titan
Moderator
Honestly, I'd suspect the motherboard. It's bad enough that you're running a 125W CPU on a bargain AM3+ motherboard that's notorious for frequently having trouble handing those at stock, but you actually sought to stress it. I wouldn't be surprised if some VRMs gave up the ghost.

And what do you mean by "eight pin to six pin?" Are you talking about the pins on the side of the cable that connects to the modular power supply? Or some kind of adapter? If you're talking about the side that connects to the modular PSU, those are proprietary and the numbers don't matter and if you're talking about adapters, you should never be using adapters with PSUs. I'm asking because it's important to have all the facts straight at the start (this is safety equipment, after all).
 
Solution
Aug 20, 2021
4
0
10
Honestly, I'd suspect the motherboard. It's bad enough that you're running a 125W CPU on a bargain AM3+ motherboard that's notorious for frequently having trouble handing those at stock, but you actually sought to stress it. I wouldn't be surprised if some VRMs gave up the ghost.

And what do you mean by "eight pin to six pin?" Are you talking about the pins on the side of the cable that connects to the modular power supply? Or some kind of adapter? If you're talking about the side that connects to the modular PSU, those are proprietary and the numbers don't matter and if you're talking about adapters, you should never be using adapters with PSUs. I'm asking because it's important to have all the facts straight at the start (this is safety equipment, after all).


Eight pin to six pin in that the first wire has one side that is 8 pin (that is connected to the PSU) and a six pin on the other side (connected to the GPU). The second wire also has one side 8 pin, but the other side is two 2 six pins. These two six pins are connected to an adapter (takes in two six pins, and outputs an 8 pin) which was then connected to the GPU. This adapter came with the PSU so I've been assuming its safe to use.

However today I took it to a pc repair place to swap out the power supplys to see if the problem persisted. Plugged in the other PSU and the same problem was happening. Turns out while it was trying to run with the case open, we noticed a burning smell coming from the GPU. So I'm assuming that the GPU had simply ran its course and finally died. Its about six years old and with the other screens graying out for a while now I suppose it makes since.
 
Eight pin to six pin in that the first wire has one side that is 8 pin (that is connected to the PSU) and a six pin on the other side (connected to the GPU). The second wire also has one side 8 pin, but the other side is two 2 six pins. These two six pins are connected to an adapter (takes in two six pins, and outputs an 8 pin) which was then connected to the GPU. This adapter came with the PSU so I've been assuming its safe to use.

However today I took it to a pc repair place to swap out the power supplys to see if the problem persisted. Plugged in the other PSU and the same problem was happening. Turns out while it was trying to run with the case open, we noticed a burning smell coming from the GPU. So I'm assuming that the GPU had simply ran its course and finally died. Its about six years old and with the other screens graying out for a while now I suppose it makes since.
Well, at least you found the problem just a shame it is the GPU as they are very expensive now.
 

DSzymborski

Titan
Moderator
OK, so these are normal cables that you're describing in an unusual manner and you used an adapter.

I'm really confused now. An EVGA G2 850W PSU has four 6+2 pin PCIE cables. Why were you using the two 6-pin connectors with an adapter? Those are rarely recommended because they're inevitably cheaply made and most people will use them on PSUs that shouldn't be powering the component in question (you are the exception as the G2 850W is still a dynamite PSU, in the good way, not the explodey way).

Unfortunately, you got just about the worst news you could get. Ideally, it would have been the decade-old stuff that died, not the good stuff that costs money. You definitely have my condolences.