Getting multiple windows hardware Errors

[Jinx_o]

Dear HWinfo Team,

A few days ago, my PSU exploded, and since then, I’ve been cautious about monitoring my system voltages. Unfortunately, I’ve encountered another issue: my Windows hardware error count keeps increasing, something I’ve never experienced before.

I’ve tried troubleshooting by removing my GPU to check if it might be the cause, but the error count still increases. Additionally, there’s a noticeable whining noise coming from my PC. While this might not be significant, I thought it was worth mentioning given the circumstances.

After further testing, I noticed something strange: the hardware error count seems to decrease with each system restart. To assist in diagnosing the issue, I’ve included eight logs in the hope that they provide some useful information.

Here’s a summary of the problems I’ve experienced so far:

1. First PSU: My original PSU (ASUS ROG Strix 1200 Aura Gold) broke down with a sound resembling a squeak.
2. Second PSU: After replacing my first psu, my second PSU worked for around 3 months. After that I suddenly heard a lot of whining/coil whine and then my PC shut down after a few minutes of operation. Shortly after, the PSU emitted a spark and made a loud bang (this was also an ASUS ROG Strix 1200 Aura Gold).
3. Third PSU: After installing a already "repaired" PSU it immeadiately startet to whine exactly the same way as the second one, that broke down minutes after.

I am now using a be quiet! Dark Power Pro 13 1300W and it seems that the only issue I am having is some slight coil whine coming from either my cpu or my AIO pump
Additional Notes:

• My GPU (ROG Strix RTX 4090 OC Edition) has always exhibited significant coil whine.
• My RAM setup (128GB DDR5, 4x 32GB) seems unstable at higher speeds.

My system specs:

• CPU: Intel i9-14900K
• GPU: ASUS ROG Strix RTX 4090 OC Edition
• Motherboard: ASUS ROG Z790 Hero
• Old PSU: ASUS ROG Strix 1200 Aura Gold
• New PSU: be quiet! Dark Power Pro 13, 1300W
• RAM: 128GB DDR5 (4x 32GB)
• Storage: 3x Samsung 990 Pro 2TB M.2 SSDs
• Cooler: ROG Ryujin III Water Cooler

Thank you very much for your time and assistance. I hope the logs I’ve provided will help identify the cause of these issues.

Best regards,
Jinx_o

 

[VictorVG]

1. A whistle and explosion of two power supplies in a row indicates a malfunction or high-voltage impulse interference in the input power supply network - it needs to be repaired and input filters must be installed on all sensitive equipment such as TV, telephones, computers, DVD players, etc. equipment, otherwise you will have to buy something else to replace the failed one.

2. Check the memory modules with the MemTest86+ test in the "All tests, one core, 4 passes" mode - it will take a long time, mine took 96 hours because this test has 4 cycles of 14 tests, the standard RAM test mode, and if the test returns an error somewhere, the module needs to be replaced. If it doesn't find any errors, then you can continue checking, but now let's exclude the RAM - its modules often break down when the power supply breaks down.

But, judging by your logs (a screenshot of the WHEA section sensors window would have been enough - everything you need was there) you have a breakdown of either the system board or in the peripheral part of the CPU circuit, since both of these nodes can become a source of PCI/PCIe bus error messages. And this breakdown occurred when the first PSU failed.

P.S.

If you don't believe me, ask any other electronics engineer to check my assumptions. I think he will fully confirm them.

At least it was easier for me - one LRDIMM out of four burned out - that was enough :

Samsung M386B4G70DM0-CMA4 Ch 1 DIMM 0 start failure, chipset freeze on to post code 0xB7 one DIMM of four: DXE Phase, Configuration Reset (reset of NVRAM settings).

I installed a spare pair of Samsung M386B4G70DM0-CMA3 - and got memory controller errors with a constant increase in the WHEA error counter, and eventually a kernel panic. I bought and installed another set - that's it, there are no errors since all the modules are the same. And it seems that Samsung M386B4G70DM0-CMA3 and Samsung M386B4G70DM0-CMA4 differ only in the iMB microcircuit in -CMA3 it is Inphi iMB02-GS02B, and in -CMA4 Montage MB CI, but that's enough.