System:
- X570 AORUS ELITE (rev. 1.0)
- ryzen 9 5950x
- 32GB RAM
- rtx 3080 SEA HAWK X 10G LHR
- Corsair HX750 (CP-9020137-EU)
TL;DR:
after months of BSOD and crashing when stressing the GPU (and sometimes, not stressing at all), managed to debug and find that the entire PSU or cables to the GPU are failing. Not sure if is the 24pin to MOBO or the 2x8 pin to GPU. most likely GPU cables/connection/entire PSU, because I changed the PSU's socket for one cable of the GPU and crashes/BSOD are gone. How should I decide if PSU needs RMA or not? I was tempted to buy a PSU tester but not sure how that works to be honest.
Long version
After months of crashes, swapping RAM sticks, trying to factory reset every possible thing on my PC, formatting, etc etc., I managed to find out that most likely either my PSU is dying, the cables are messed up, or the GPU sockets are failing/not connecting correctly (really doubt this last one).
the decisive test that concluded that the failing point is between the GPU and the PSU, was when I decided to change one of the PSU output sockets for the GPU (I have 2x8 pin cable comming from the PSU to the GPU), and the issues were no more. After that I started to look more into the values and possible readings on the PSU related values, and I came up to this (eyes on the 12V and 5V rail):
Should I try doing readings with a different software? is this looking bad? I was also thinking of buying a PSU tester to confirm the results instead of going around with a multimeter (since I don't know how to do the readings, I can only handle very small and basic circuits when it comes to that).
Should I just purchase a new PSU and test the values with a new one to see how it goes? really need some help here, since I would like to have this system running for at least another 2~3 years
thank you for the help
- X570 AORUS ELITE (rev. 1.0)
- ryzen 9 5950x
- 32GB RAM
- rtx 3080 SEA HAWK X 10G LHR
- Corsair HX750 (CP-9020137-EU)
TL;DR:
after months of BSOD and crashing when stressing the GPU (and sometimes, not stressing at all), managed to debug and find that the entire PSU or cables to the GPU are failing. Not sure if is the 24pin to MOBO or the 2x8 pin to GPU. most likely GPU cables/connection/entire PSU, because I changed the PSU's socket for one cable of the GPU and crashes/BSOD are gone. How should I decide if PSU needs RMA or not? I was tempted to buy a PSU tester but not sure how that works to be honest.
Long version
After months of crashes, swapping RAM sticks, trying to factory reset every possible thing on my PC, formatting, etc etc., I managed to find out that most likely either my PSU is dying, the cables are messed up, or the GPU sockets are failing/not connecting correctly (really doubt this last one).
the decisive test that concluded that the failing point is between the GPU and the PSU, was when I decided to change one of the PSU output sockets for the GPU (I have 2x8 pin cable comming from the PSU to the GPU), and the issues were no more. After that I started to look more into the values and possible readings on the PSU related values, and I came up to this (eyes on the 12V and 5V rail):
Should I try doing readings with a different software? is this looking bad? I was also thinking of buying a PSU tester to confirm the results instead of going around with a multimeter (since I don't know how to do the readings, I can only handle very small and basic circuits when it comes to that).
Should I just purchase a new PSU and test the values with a new one to see how it goes? really need some help here, since I would like to have this system running for at least another 2~3 years
thank you for the help
