Power issues continue even after replacing the PSU, I'm confused!
Power issues continue even after replacing the PSU, I'm confused!
I’ll review a list of actions I’d take in a similar situation, regardless of order, but I’ll start with the simplest steps first.
1) Check the temperatures of all components—m.2 SSD temps, CPU temps, GPU temps, motherboard temps, since any of these can lead to unexpected shutdowns.
2) Look at your RAM settings; this should come near the top of my list. With four sticks and a 5800x, you might need to set it to 3200mhz if running higher speeds, or even lower if performance suffers. Running with two sticks could ease the load on the memory controller.
3) Reset the BIOS to factory settings, skip tweaking XMP/DOCP unless stability improves, and test RAM speed changes.
4) Update the BIOS, though it’s unlikely, a fresh update can sometimes improve stability—avoid beta versions.
5) As a last option, reinstall Windows and try a different drive to see if stability returns after installation.
6) If things go wrong, you might encounter a faulty 5800x with memory issues; replacing the CPU could resolve the problem.
Good luck!
Thanks for your help! Temperatures look acceptable, CPU is idling between 30-40 degrees, GPU and m.2 SSD are stable, and idle temps are normal. In tests where they got hot, everything remained under control and no overheating crashes occurred. The Windows reinstall and BIOS update were completed successfully, but no improvement was noticed in the crashes. I assume they were always set to defaults; I’ll check the RAM speeds next.
I’ve owned a 5800x for three years now, and if this issue has recently appeared, it’s really disappointing.
I plan to test with just two RAM sticks in the second and fourth slots to see how it performs, thank you!
I also updated my original post since I’ve gathered more details from HWiNFO and Event Viewer during my OCCT tests.
Edit: I’m currently running a Power test again (CPU & GPU) and it’s lasted six minutes so far without a crash (previously it would crash after 3 seconds). I’m not sure what changed, but it’s still showing WHEA errors during the test (6 minutes in).
Ram is an unusual component, especially when unstable, as it can cause crashes in Windows applications, giving the impression of a program or file being involved. It can also lead to video driver failures. I recently encountered a situation where I believed an SSD was failing, but despite showing write activity, data wasn't being saved—this happened even with a RAM stick. That incident was quite amusing, to say the least.
I remain convinced it’s related to RAM or memory controller problems, given that these are 16GB modules. Running one at a time might help identify whether errors persist or reappear based on the specific stick or channel.
This is something I’ve dealt with frequently over the past 20 years of working on PCs—memory/RAM issues always seem to mask the real cause, making it hard to pinpoint the exact problem.
Hopefully it's just a matter of unstable RAM, I'm not certain why it started suddenly without any changes, though! It seems like BIOS and Windows updates might have contributed somehow? I’d really like to set up the RAM so these crashes stop, rather than replacing it, which I’m open to doing but prefer to avoid.
I ran another CPU test, which this time worked and returned just over 200 WHEA errors, all in the "CPU Cache L1" area. From what I’ve read, this could be linked to undervolting. I’ve left everything at default settings, but I’ll try adjusting it to see if it helps. It doesn’t seem to be the main cause of the crash, though I’m not entirely sure.
Thank you again for your help and support—it means a lot. I appreciate it and hope we can figure this out together! I’ll let others know if they encounter similar problems and are looking for a fix. I’m also happy to answer any other troubleshooting questions that might help identify the issue.
Have you performed a detailed RAM check using MemTest86+ from a USB drive? With 64GB, a full assessment might require 3 to 4 hours, though it’s worth the effort to verify the integrity of your RAM.
https://memtest.org/
If MemTest86+ reports any issues, lower the XMP overclock setting. Should the RAM still fail at the standard JEDEC 2133MT/s, discard the problematic DIMM(s). When testing all four modules simultaneously, consider rechecking each stick individually to pinpoint the faulty unit. If you adjust the XMP speed, rerun MemTest86+.
I own four Vengeance DDR4-3000 modules in an older 3800X system and they function reliably at 3000MT/s, though some users on this forum mention issues with Vengeance and Ryzens.
RAM defects can emerge unexpectedly if they experienced ESD (static) damage during installation or removal, sometimes not showing symptoms immediately but potentially lasting months or years.
If the problem lies with RAM it usually leads to Windows corruption and may produce unusual symptoms that eventually cause crashes. There is a scan that can be executed in an admin elevated CMD, which checks for this kind of corruption and tries to resolve it. If your RAM is faulty, the chances that this scan detects corruption are nearly 100%.
1. Open CMD with administrative privileges by right-clicking the shortcut.
2. Enter or copy the command: "sfc /scannow" and press enter.
If the scan shows no issues, it strongly suggests your RAM is intact, although a clean scan with bad RAM is possible. The process takes roughly 2 to 5 minutes.
As @geofelt mentioned, this could still point to a software or driver problem rather than a hardware failure. It’s recommended to perform a "clean boot" of Windows as described, to eliminate software-related issues. After testing, run your checks again and revert any changes if needed. Please let me know your results after these tests!
I've run a MemTest86+ test with all four drives and everything has been perfectly normal so far. However, I haven't tested them individually, just to be safe. According to what I know, everything is set to default or auto in the BIOS, so there shouldn’t be any XMP-related tweaks needed.
Unfortunately, it’s difficult to identify the problem since the crashes have been happening for a while but were very infrequent until last week. Apart from a Windows 11 update from Windows 10 and a BIOS update, nothing has changed in terms of hardware. A fresh Windows installation also didn’t resolve the issue.
Before I started a clean Windows reinstall, there were definitely corruption problems. I recall running sfc and dism checks, but the computer would crash during those processes. On one occasion, it finished an sfc check, but there were corruptions and it couldn’t fix them. Dism also failed, reporting that the component store was damaged. After a clean Windows reinstall, both sfc and dism completed successfully, yet the problem remained. I can also confirm it still occurs during a clean boot, which is frustrating—it kept happening eventually, randomly.
The only methods to encounter such corruption with a verified good RAM setup are if it's happening on another memory layer or if data from various parts isn't being properly received. This points to potential problems with the CPU, OS drive, and motherboard. It seems like just one of these hardware components might be faulty. I’d probably guess the CPU is involved due to the nature of the errors. The best way to confirm would be to replace suspected parts and observe if the issue resolves. If others have better ideas, I’m listening.
I really plan to try another CPU next. I’ll check if I can get one or if it’s worth purchasing for testing.