PC stops working unexpectedly, restarts briefly to turn on again and then powers off immediately.
PC stops working unexpectedly, restarts briefly to turn on again and then powers off immediately.
The PC has been experiencing some problems. Recently upgraded, it seems that on certain days when trying to power it on through the case power button, nothing happens—no lights, no fans, nothing turns on. Reconnecting the PSU cable connected to the surge protector power strip resolved the issue, allowing the computer to boot normally. Earlier today, after logging into the Windows desktop, the PC shut down unexpectedly. When attempting to restart it via the case power button, lights and fans briefly activated before shutting off again. Pressing the power button continuously tried to turn it on. All other connections were removed except the PSU cable, which showed no changes. The power strip remained stable, not tripping or blowing a fuse. All cables were inspected and properly seated. I replaced the power strip with another identical one used elsewhere, maintaining only the PSU cord; the PC started without issues. I then reconnected the monitors, and the PC remained on for about 30 minutes before shutting down abruptly. Could this indicate a PSU problem or an issue with the CPU/Motherboard? I haven’t tested the PSU with a multimeter yet, but I’m hoping someone has already found an answer. Using the PSU power cord provided with the unit, temperatures on the CPU reached 42°C. Any advice would be greatly appreciated.
Plug the PSU straight into the wall and see how it performs. Also, consider letting it run for a while. Many modern motherboards need several attempts to adjust memory timings and other CMOS settings properly. I've observed systems repeatedly powering on, shutting down, and restarting up to six or seven times before settling into their settings. After a successful POST and when everything is set correctly, it's wise to enable the memory fast boot option in the BIOS under advanced memory configuration. This prevents it from re-training itself during cold starts or power loss. Of course, this doesn't affect the behavior when you press the power button. First, remove the power strip entirely and, if using a UPS, eliminate that as well. If it works without any of these, proceed to thoroughly test the PSU with a multimeter.
So sadly I didn't get the chance this evening to further test the PC by connecting it directly to the wall. I know a PC can restart several times during training setups, but I hadn't made any changes or replaced any parts since it was finished about three months ago. This weekend I managed to turn it on to check if the issue continued, and on my first try at starting everything, it booted up perfectly. It ran for six hours without shutting down or cutting power. I took out the PSU and checked it with my multimeter—it was fine, no problems—but it seems there was no load connected. No cables were running under the motherboard to cause a short, though some people say standoff screws might be the culprit. Still, it hasn't happened before, and I'm puzzled by this recurring problem. The Event Viewer shows only some TPM errors with a fast boot error code 0xC00000D4, which doesn't seem connected to the sudden power loss. After the PC briefly flicked on and then turned off, the PSU didn't need resetting either, so I'm more inclined to think the motherboard is the problem. I'm running out of options since I don't want to send it back for repair and then wait for another shipment.
I was quite upset because I struggled to reproduce the problem or understand what was happening. After reaching out to EVGA support, they suggested running OCCT and stress-testing the PC for ten minutes. I relocated the entire system to another room with a different power outlet, and it booted normally. I verified temperatures via the BIOS—CPU at 38°C, GPU at 42°C—and confirmed the 12V rail stayed steady at 12.153V without variation over five minutes. I downloaded OCCT and, just before testing, the PC shut down again. After restarting it for ten minutes, I ran the test once more; the PC remained stable, with temperatures holding at 69°C and GPU at 89°C. I shared all the details with EVGA, who approved an RMA. When I spoke to the physical store where I bought the PSU, they were able to replace it directly for me instead of shipping. They only had a 1300W SuperNOVA G+ unit in stock, so I paid the difference. I also contacted ASUS for an RMA regarding the motherboard, though my serial number was listed as overseas. Since I bought the board through MemoryC, I reached out for an RMA there. In the meantime, I removed the motherboard from its case to test if the CPU fans would spin during boot attempts, and it worked fine. I also swapped the wall power outlet for the PC and monitors. I sent this information to EVGA, who confirmed the repair. I’ll update this post within a week at the latest, or sooner if needed, and provide an RMA update for the motherboard or PSU. Thank you all for your assistance and the guidance—hopefully this issue is resolved.
Following the installation of new parts, the PSU and motherboard still showed the same problems after a week of consistent operation. Someone suggested that Kernel Error 41 might stem from incompatible universal RGB control software. Eventually, after restarting into Windows and removing the Patriot and Holtek drivers, the PC failed within two minutes. I tried using a separate SSD with a fresh Windows installation, but the computer still crashed upon startup and would flicker regardless of changes. A local repair shop tested each part individually on a dedicated bench and confirmed the CPU was defective. The B550 motherboard displayed identical symptoms. All other components seemed normal, though the same company that handled my PSU RMA had previously identified it as faulty, with failed 12V rails. A new CPU will be ordered, and I hope this resolves the remaining issues. I’ll also request a replacement PSU, as the 1000W/1300W G+ / G1 models are considered low-tier on the forum’s power supply list.
Confirm if it is indeed the CPU. That would, in my opinion, indicate a significantly higher failure rate than expected for relatively new AMD chips compared to older models. I notice many discussions about failures in AMD processors that shouldn't occur, especially over the past couple of years.
Sorry to bring this up again, but I hadn’t updated or closed it before. Problems came from a defective 12V rail on the EVGA G+ 1000W PSU, along with a faulty 5950X and a failed part on the ASUS ROG X570 Crosshair VIII Formula. After filing an RMA for the 5950X, motherboard issues disappeared for about a year and a half. Then I experienced similar shutdowns, but this time it was due to the Corsair iCUE software. Removing that caused no further problems.