Question PC shut off randomly, when attempting to turn back on powers on for less than a second and then shuts back down.

[Riptide-]

As title suggests, PC has been having some issues. Over the past few months after upgrading my PC, it would seem some days when attempting to turn on PC through case power button, nothing would occur. No lights, no fans, nothing would turn on. I would unplug PSU cable going to surge protector power strip (Phillips SPC3064BG/37) plug it back in, and computer would boot normally. Earlier today, starting PC went okay, after getting into Windows desktop after signing in, computer shut down without warning. Attempted to power PC back on with case power button and lights/fans would turn on for a split second and then shut back off. Holding the power button would just continue to attempt to turn it on. Removed all other connections going to power strip except PSU cord, no changes. Power strip never tripped nor did breaker. Removed all cables from PC and re-checked all wires on PSU and all components were seated properly. All seemed to be properly seated. Switched out power strip with identical one that was being used elsewhere in home, same setup of only PSU cord being plugged in and PC booted up no issues. Plugged monitors back in and PC stayed on for about 30 minutes and then shut back down without warning. Could this be a PSU issue or a CPU/Motherboard issue? I haven't tested the PSU yet with my multimeter but wished to see if anyone has an answer already. Using PSU power cord that came with PSU. Temps while on for CPU are at 42°C. Any help is appreciated.

PC specs: https://pcpartpicker.com/list/8zcX2m

 

[Darkbreeze]

Plug the PSU directly into the wall and try it.

Also, have you let it just go for a while? Many modern motherboards require going through several, in some cases MANY, training attempts before they are able to find suitable settings for memory timings and other CMOS settings. I've seen systems power up, shut down, and power back up, as many as six or seven times before they were satisfied with their training configurations. And, once you have a successful POST, and everything is configured the way you want it, it's not a bad idea to go into the BIOS and enable the memory fast boot setting in the advanced memory configuration so that it doesn't try to re-train every time it cold boots or loses power.

Obviously that won't have anything to do with it not doing anything when you hit the power button though. I would first completely eliminate the power strip and if you are using a UPS battery backup, I'd eliminate that as well. If it does the same thing plugged directly into the wall, then I'd pull it all out and bench test it with the most minimal hardware possible, as outlined here:

How To - Motherboard bench testing

Finding the problem through bench testing If you are here then it’s likely you have encountered a serious hardware issue and have been unable to resolve it using the standard no-POST troubleshooting procedures. If you have not yet attempted to resolve your issues using the no-POST...

forums.tomshardware.com

And if it STILL does it then, I'd probably recommend that you think about an RMA on the motherboard. Before bench testing though, it's probably a good idea to test the PSU with a multimeter.

View: https://www.youtube.com/watch?v=ac7YMUcMjbw

 

[Riptide-]

Plug the PSU directly into the wall and try it.

Also, have you let it just go for a while? Many modern motherboards require going through several, in some cases MANY, training attempts before they are able to find suitable settings for memory timings and other CMOS settings. I've seen systems power up, shut down, and power back up, as many as six or seven times before they were satisfied with their training configurations. And, once you have a successful POST, and everything is configured the way you want it, it's not a bad idea to go into the BIOS and enable the memory fast boot setting in the advanced memory configuration so that it doesn't try to re-train every time it cold boots or loses power.

Obviously that won't have anything to do with it not doing anything when you hit the power button though. I would first completely eliminate the power strip and if you are using a UPS battery backup, I'd eliminate that as well. If it does the same thing plugged directly into the wall, then I'd pull it all out and bench test it with the most minimal hardware possible, as outlined here:

How To - Motherboard bench testing

Finding the problem through bench testing If you are here then it’s likely you have encountered a serious hardware issue and have been unable to resolve it using the standard no-POST troubleshooting procedures. If you have not yet attempted to resolve your issues using the no-POST...

forums.tomshardware.com

And if it STILL does it then, I'd probably recommend that you think about an RMA on the motherboard. Before bench testing though, it's probably a good idea to test the PSU with a multimeter.

View: https://www.youtube.com/watch?v=ac7YMUcMjbw

So unfortunately I didn't have time the rest of the evening to test the PC any further by plugging it in directly into the wall. I do understand that a PC will restart multiple times in training configs, however I had not changed any parts/components since it was completed about 3 months ago. This weekend I had a chance to turn it on to see if the problem persisted and on first attempt at starting, everything booted up just fine. Let it run for 6 hours and it did not shut down/power cut itself at all. I removed the PSU and checked it with my multimeter and tested okay, no issues but obviously there was no load on it. No cables are ran under the motherboard to cause a short but some say standoff screws can cause the MoBo to short. However it hasn't happened since and I don't know what's causing this annoying intermittent issue. Event viewer also doesn't seem to log anything besides some TPMs errors alongside a fast boot error with code 0xC00000D4 but that doesn't seem related to the sudden power loss. After the PC would flick on for a split second and turn back off, the PSU didn't have to be reset either, so I'm leaning more into the MoBo being the issue here. I'm running out of things to test since I don't want to RMA the wrong thing and then have to wait for shipping all over again.

 

[Darkbreeze]

so I'm leaning more into the MoBo being the issue here. I'm running out of things to test since I don't want to RMA the wrong thing and then have to wait for shipping all over again.

Which is why I said

And if it STILL does it then, I'd probably recommend that you think about an RMA on the motherboard. Before bench testing though, it's probably a good idea to test the PSU with a multimeter.

 

[Riptide-]

Which is why I said

I'm very sorry, I was just frustrated with the issue as I couldn't replicate it easily or figure out what was going on. After contacting EVGA support, they advised to run OCCT and run the PC with a load for ten minutes. I was able to move the entire PC over to another room with a different outlet, and everything booted fine. I checked temps for the CPU and GPU through BIOS and checked the 12V rail voltage as well, sitting at 38°C for the CPU, 42°C for the GPU and the voltage was sitting at 12.153V consistently with no fluctuations over the span of five minutes. I downloaded OCCT, and just as I was going to test it, the PC shut down again. I was able to turn it back on and run the OCCT test for ten minutes, then ran it again for another ten. PC never shut down, temps stayed at 69°C and GPU was at 89°C . I sent the info to them, I also stated I would leave it running overnight to see if it would shut down again. Left it running from midnight to about 10 A.M. when I woke up. I found the PC to have shut down again, and attempting to start it, I finally had recreated the issue and the PC would flicker on for a split second and turn off. I then attempted to remove the front tempered glass panel to see if the CPU fans would spin while attempting to boot, and then the PC booted fine. I then also relayed this information to EVGA, they approved an RMA. However when contacting the brick and mortar store I bought the PSU from to receive a copy of the receipt, they were able to swap the PSU for me directly so I wouldn't have to ship it. All they had in stock was 1300W SuperNOVA G+ so I payed the difference. I also contacted ASUS for an RMA on the motherboard, however my serial number on the motherboard was an overseas serial number. I had purchased the motherboard through MemoryC so I have contacted them for an RMA. In the meantime I didn't want to wait and had already removed the motherboard from the case so I purchased a replacement motherboard. Everything has been reinstalled with the replacement motherboard and replacement PSU. I also replaced the wall receptacle that was originally powering the power strip to the PC and monitors, and moved everything back over. I will update this post in a week at the latest if not sooner if the problem arises again and/or an update on the RMA'd motherboard/PSU. Again, thank you for your help and all the tips and advice on what to check, hopefully my issue is resolved.

 

[Darkbreeze]

 

[Riptide-]

After installing the new components, PSU and motherboard, PC exhibited the same issues after one week of stable use. Found advice that Kernel Error 41 could be caused by conflicting universal RGB control software. After eventually booting into Windows and uninstalling Patriot and Holtek universal drivers, PC crashed within two minutes afterwards. Attempted to use separate SSD with clean install of Windows and computer crashed while booting. Computer would no longer turn on and would flicker no matter what was changed out. Had a local computer repair company test every individual component on a separate test bench and the results were that the CPU was the faulty component. Thrown on a B550 motherboard, it showed the same exact symptoms. All other components are fine, however this company was the same company that I had RMA'd the PSU with. They had tested the PSU themselves before sending it back to EVGA and had found that the PSU was also faulty and that the 12V rail failed their tests. New CPU will be ordered and hopefully all problems will be resolved. Will also order new PSU as the 1000W/1300W G+ / G1 power supplies are low tier PSU's on the forums tier list.

 

[Darkbreeze]

Confirm, if it does turn out to be the CPU. That would, in my eyes, signal a rather higher than normal failure rate on relatively new AMD processors as compared to any kind of processor from previous generations. And I say that because I've seen quite a lot of threads involving failures on AMD processors that shouldn't be faulty, over the last two years or so.