Computer randomly shutting down

rads2314 · June 3, 2021

ryzen 3700x

Asus TUF b550M wifi

Corsair Vengeance LPX 16gb 2x8 DDR4-3600 cl18

SN750 500gb m.2

RTX 3060 12gb / gtx 660

Fractal Meshify C

Fractal ION+ 660w platinum / Corsair rm750x (2021)

I tested this system first with an RTX 3060, then with a gtx 660, I still get the computer to randomly shutdown. As well as with testing two different power supplies, the problem still persists.

No warning before shutting down. Windows event viewer just says Critical: Event 41 Kernel-Power: "The system has rebooted without cleanly shutting down first". (The bug has all zeros)

I don't know, maybe the cpu is sucking power at an extremely high rate randomly? Is there something wrong with the motherboard, memory, or SSD? Those are the parts left I guess by process of elimination. I say random because I could use the computer fine and run a game for 7+ hours, so it's hard to replicate a moment when it might shutdown, toubleshooting is hard to test when so much time has to elapse to see if the computer might shutdown again.

edit: I don't know, maybe its a power deviation thing that HWINFO64 shows. The crashes do always happen when the computer sits at idle. I think the power deviation has gotten really low before to like 50%. But searching online people say the power deviation is irrelevant. However it is talked about that power deviation represents some kind of communication between the motherboard and cpu. Oh yea and one final observation, I thought this was anecdotal and could throw off any investigating, but the shutdowns did appear much more frequently and violent when at first, it was the Fractal 660W PSU and RTX 3060 (lower wattage supply and higher consuming components), and less frequent to where I am now, using a 750w supply and only a gtx 660. Again it seems anecdotal, but I'm guessing the next step I should take is reseating the cpu? I'm guessing I will only end up having to repurchase another one. Does anyone know if there is some extreme underclocking I could do to test this out to see if the CPU isn't acting like a vampire to my power supply?

Mark Kaine · June 3, 2021

Nothing sounds really like a cpu issue, but u can try reseating i guess.

how are temps under load typically?

generally, this is more likely psu, motherboard or bios related. so try updating the bios *and* chipset drivers.

edit: also max cpu voltage in hwinfo64 ? and whats the voltage at while idle?

rads2314 · June 3, 2021

On 6/3/2021 at 3:17 PM, Mark Kaine said:

Nothing sounds really like a cpu issue, but u can try reseating i guess.

how are temps under load typically?

generally, this is more likely psu, motherboard or bios related. so try updating the bios *and* chipset drivers.

Temps are normal. I had the borderlands 3 menu open for 7 hours and got 50C cpu average and 61C gpu average. In game it was the same result, fps was fine. Used the highest quality setting 'badass', their built in benchmark was also the same expected results. It can't be the psu or gpu in my opinion. Right now I plugged that fractal psu and the 3060 into my decade old computer with an i5 3570k and asrock z77 extreme 4, no problems yet. I updated the bios firmware, that alone wasn't a fix I remember. But I do admit I didn't realize I was also supposed to do their long list of drivers & tools, so thats embarrassing on my part, I'll try and get results after that. Other than that my idea was to default clock the cpu because every setting in the bios is set to auto.