Device turns off automatically without intervention
Device turns off automatically without intervention
After extensive investigation, I haven't discovered any clear explanation or fix for these problems. Most experts likely have better insights. The main concerns are twofold, primarily one: A) My Ethernet card fails after several hours—sometimes around 4 hours, other times up to 6. It remains disabled and can't be reactivated. Before this point, I could toggle it on or off without issues. Once I switched to Wi-Fi and connected to the same network, the connection lasted longer. The issue appeared consistently in both adapters, though it seemed to recover after some time, possibly due to the Ethernet card restarting intermittently. Temperatures stayed below 70°C, and a full PC reset resolved everything. B) A recurring black screen followed by GPU fans ramping up to maximum speed, stopping only after a manual restart. This happened twice and might be connected. I've seen error codes like 47, 41 (after startup stays 41), 49, 45, 42. All drivers were cleaned and reinstalled from Asus directly. The NVMe drive was previously in a laptop, so I swapped it out and cleaned everything thoroughly—stress tests on memory, CPU, GPU, and PSU showed no faults. The PSU cables reached the power supply but weren't overly loose; they weren't exposed to harsh conditions. The case is fully updated with Windows 10 Home x64, AMD Ryzen 7 9800X3D, and the graphics card is a high-end model. If you have any suggestions or solutions, I'd appreciate it!
That sounds like NV driver plague. Since the 50 series launch they've been terrible Honestly, that's probably just a dying NIC. It should qualify for RMA, but TBH, when faced with the option of a mobo swap (which in my experience takes longer than building a whole PC though YMMV) or shelling out for a 20$ NIC (or 60$ for a 10G NIC), I'm pretty much always gonna go for the NIC. I was gonna make a caveat about SFF PCs, but it seems that while you have a SFX PSU, you have a full sized mobo. There is a chance that transplanting the Windows install to new hardware messed something up. First you should definitely do runs of SFC and DISM and check for any microcode optional updates. If that doesn't fix anything, one option would be to try booting a linux LiveUSB and see if you get the same problems on that. If so, it's definitely a hardware problem. If not, you may be in luck and only need to reinstall windows.
Just to be safe, NIC is network interface card, right? And I believe so too, at least by the looks. I can claim RMA asap and swap it for the same in like a week (or any MB better in that fact), but im doing some hopefully a last test next week swapping the MB for the 'ASUS ROG Strix X870E-E Gaming' a little overkill but should give me answers. For the PSU i saw that it was powerful and reliable enough even for SFX to beat my other options non SFX, so far it goes golden, only like I said cables are not the longest but maybe cablemod in the future.. I forgot to say I did SFC and DISM already, but I did it again just in case. The linux idea is great, I'll try to test with the upcoming x870 first and then linux, as I have also the potential doubt that could be the cpu not sitting properly and that might show why all those codes, now btw showing 40, 41, 42. Cause from these codes comes to my mind bad sitting cpu, faulty ram, faulty cpu or wrong bios ram config. But I believe the ram works flawless and sits fine, and the cpu is fine but maybe not sitting properly, still will not make the network adapter disable itself just randomly for what I know. Yesterday I had the issue once again, here is what I did or happened. I was watching twitch while having a call in discord and friend told me to play Star Citizen, updated the game started playing, like 2hrs in, ethernet disables itself, I go wifi, after like 4 hrs? The wifi adapter still working but shows as it can't find any network (on but not showing) I run win troubleshoot on wifi and immediately finds and 'resets' itself and shows me all networks and auto connects, funny is that it restarted also the ethernet and I got both adapters up and running again I run the troubleshoot again on the working ehternet again and shows as 'ethernet does not have a valid IP address' but I'm connected and working in ethernet. To mention that I run cloudflare 1.1.1.1, but it causes no issues in my eyes.
This situation isn’t caused by a basic driver error. Because both Ethernet and Wi-Fi eventually stop working and only resume after a restart or troubleshooting, the issue likely lies with the chipset or firmware. I’d run a Linux live USB and keep it running for a few hours to check if the problem persists. If it does, the hardware is faulty. If not, it’s probably related to Windows or BIOS settings. You might also attempt to disable C-States or ASPM in the BIOS. If the behavior remains unchanged, swapping to the X870 board will help determine if your current motherboard is the cause.
I think the issue might be on the chipset or NIC itself. I'm using the 3.4.2.1046 driver for the MediaTek Wireless LAN card (mediatek wifi 6e mt7922 rz616), which seems to be the cause. Interestingly, Asus drivers appear stable, but I'm still seeing codes like 40/41/42. I looked up the specific NIC and found another similar problem, and discovered a newer driver version 3.4.0.1268 from CCleaner. I also found a library with driver 3.4.0.1123, but it's unclear how to install it manually.
Recent checks on those drivers show the NIC crashes have dropped markedly, occurring only once so far (today and possibly again after updates, but much less). There are still noticeable 'spikes' where the network seems unstable. I received the MB earlier, but more tests are expected soon.
I experienced issues with my ASUS TUF B650 using the NIC drivers and Windows 11. I don’t remember the previous version, but updating the drivers from the official site fixed the problem about six months ago.