CyberpowerPC motherboard or CPU issue?

[SpookyCitrus]

So I'm working on a friends computer, it's a CyberpowerPC with a Ryzen 5 3600, Gigabyte B450M DS3H WiFi, 16GB 3000mhz Crucial Ballistix memory, and a MSI Ventus GTX 1660. It started to not post and the PSU which is the EVGA GD Gold 600w was making a double click every time you try to power it on and will also come on for 2 seconds and then power off immediately repeatedly just on off, on off, over and over while double clicking. He thought it was his PSU so he replaced it, it now boots into Windows but will randomly restart without warning and without a BSOD or any sort of log or error message. I have tested the old PSU that was having the issues with a both a PSU tester and by hooking it up in another gaming rig and it worked fine with both so I'm positive it is not the PSU. So far these are the steps I have done.

1. Bios update to the latest available.
2. All drivers and chipset updates.
3. Reseated the CPU.
4. Ran Memtest86 (Restarts randomly while running the test, no memory errors in logs)
5. Tested in Winpe (Also restarts randomly)
6. Removed motherboard and components from case and have them on a motherboard box for bench testing. Issues persist all the same as normal however the EVGA PSU that would just click and power on and off is now posting and starting to go into Windows before it clicks and powers off and on again instead of every few seconds.

As of right now I'm leaning towards the motherboard or CPU being the issue, as looking through other forums and posts with similar issues have had their issue fixed by replacing one or both of them. My next step is to test with another CPU to see if the issue still persists with a completely different CPU.

Please let me know if you have had a similar issue or have a suggestion.

[PorkishPig]

26 minutes ago, SpookyCitrus said:

• Ran Memtest86 (Restarts randomly while running the test, no memory errors in logs)

If Memtest86 is failing while running tests, it's probably memory. Reseat the memory, booting with each stick individually.

[SpookyCitrus]

11 minutes ago, PorkishPig said:

If Memtest86 is failing while running tests, it's probably memory. Reseat the memory, booting with each stick individually.

I have already done that, also tested it in another computer. Memtest is not failing and it is not crashing, the computer just randomly restarts, it can be after 1 full pass or two or after one second of booting into it. Also checking the logs, which log if there is an error that causes memtest to crash there is none. I have tested with completely different memory and the issue persists. I have just tested with a different CPU as well and it is still doing it. The only item I do not have to test is another motherboard, I believe it is the culprit. Thanks for the suggestion.

[ZephCloud]

Are the 2 pc plugged into the same wall socket when doing the test?

[SpookyCitrus]

5 minutes ago, ZephCloud said:

Are the 2 pc plugged into the same wall socket when doing the test?

What two pcs?

[PorkishPig]

16 minutes ago, SpookyCitrus said:

I have already done that, also tested it in another computer. Memtest is not failing and it is not crashing, the computer just randomly restarts, it can be after 1 full pass or two or after one second of booting into it. Also checking the logs, which log if there is an error that causes memtest to crash there is none. I have tested with completely different memory and the issue persists. I have just tested with a different CPU as well and it is still doing it. The only item I do not have to test is another motherboard, I believe it is the culprit. Thanks for the suggestion.

Gotcha, I figured that it was turning off suddenly during tests after entering a kernel panic and failing to write to the logs.

 

I only have experience with a similar issue in the enterprise space. The server would boot normally, but after a few minutes it would just shutdown/restart unexpectedly. It ended up being improper cooling, where the board was trying to indicate that there were fan headers that needed to be populated to be stable. Seems unlikely to be the culprit, but check to make sure that both CPU_FAN and SYS_FAN1 are filled. Otherwise, it sounds like a dead motherboard.

[SpookyCitrus]

1 minute ago, PorkishPig said:

Gotcha, I figured that because it was turning off suddenly during tests after entering a kernel panic and failing to write to the logs.

 

I only have experience with a similar issue in the enterprise space. The server would boot normally, but after a few minutes it would just shutdown/restart unexpectedly. It ended up being improper cooling, where the board was trying to indicate that there were fan headers that needed to be populated to boot. Seems unlikely to be the culprit, but check to make sure that both CPU_FAN and SYS_FAN1 are filled. Otherwise, it sounds like a dead motherboard.

The CPU fan is hooked up correctly in the CPU_Fan header whenever it is powered on. I have never had a consumer level motherboard need the Sys_fan1 header populated to boot, I can see a server but not a standard consumer board. But yes I have tested it with the fan headers populated and not, it has the same issue no matter what, it's definitely a power issue with the board.