PSU issues vary unpredictably (potential hardware or power supply fault)
PSU issues vary unpredictably (potential hardware or power supply fault)
Observed running demanding games or GPU benchmarks would cause computer to turn off. You can hear an audible click come from the power supply and the computer remains off. Only way to turn it back on is to switch off PSU, wait 30 seconds, switch on PSU and press the power button. I can get this issue to happen repeatedly by running the GPU Compute test on Passmark’s Performance Test. Happens almost immediately when the second test runs. I thought this was due to running the GPU with an OC, so I dialed it back to stock settings but the issue remained. I borrowed a buddy’s RTX 3080, ran the same test at the highest power and OC settings I could in MSI Afterburner and each test completed successfully. I took out my main GPU (RX 6750 XT) and ran off of the iGPU. Left the computer on overnight and found the computer turned off. Sure enough, pressing the power button did nothing. Had to switch off power supply to get the computer to turn back on. At the same time, I have had issues with this motherboard ever since it was bought as a combo deal at MicroCenter. I purchased an AMD Ryzen 7 7700X + ASUS ROG Strix B650E-F kit and had two issues since day 1. 1) The computer would do the same PSU trip when rebooting. This happens 50/50 when performing a normal restart. 2) The memory training takes forever, even with turning the memory retraining feature in BIOS off. ASUS assured me that the issue would be fixed in future BIOS revisions, but that was a year ago and no different in boot times have been observed since. I have tried: - Ran DDU and reinstalled GPU drivers - Flashing BIOS - Reseating every component and cable - Removing all parts from tower and running on my MoBo box testbench - Toggling each stick of RAM, trying different slots and number of sticks. - I don't have another MoBo or PSU to test. Given the issues of the MoBo since day 1 and the reputation I believe Seasonic PSUs to have, my belief is that it is the MoBo that is the problem. Windows 11 64 bit ASUS ROG Strix B650E-F (running latest BIOS from 02/2024) AMD Ryzen 7 7800X3D (all issues described above were observed with the 7700X and now this 7800X3D) 2x32GB G.Skill Trident Z5 RGB (all issues described above happened when running the AMD RAM OC profile 1, tweaked and when running and JEDEC) PowerColor Red Devil RX 6750 XT (had the card OC’d to 2800MHz for the longest time, but now won’t run well at 2600MHz) Boot SSD is a Samsung 980 500GB m.2 Seasonic TX-750 PSU
I would have swapped the PSU already. #2 isn't connected. This should be addressed in another thread—usually making sure Memory Context Restore is active. (unless it's not training memory and booting just a bit longer, like my X670E Taichi which takes about 30-35 seconds from press to desktop). This is probably the PSU; anyone who does this knows you can't rely on brands, you rely on models. Even then, components might fail. You must test it, and the only thing left is swapping parts. Something definitely triggers one of the protection circuits.
Thank you for the input, it’s clear I’m conscious of this bias. Without additional PSUs or MoBOs to evaluate, it looks like returning the PSU would be the most practical option. It would be inconvenient to purchase or send a PSU from my local Micro Center, and I’d rather avoid paying for their assessment while my computer remains unavailable.
The fact that the PSU powers on and off suggests a problem with the power supply unit itself. The Seasonic Prime TX has experienced shutdowns when using certain GPUs. Disconnecting the 24-pin sensor wire might resolve the issue, or you could replace it with a better non-Seasonic PSU.