Computer Randomly Shuts Off

whiteguytyrone

Prominent
Sep 13, 2017
2
0
510
I've had this problem for a while now, while I'm playing a game or just casually using my computer it will shut off without warning. I know this is a power supply issue, however I have tested the voltage and they all read correctly. Immediately after a recent crash I learned the power supply just stops working and doesn't even power the tester when it crashes. There seems to be no heat issue since it doesn't feel hot, definitely not an overload since I calculated my estimated voltage and made sure I had a large enough power supply. Even when I start the computer it will start for a second, shut off, and then start up again.

What I'm here for is I want to know why this is happening, this is probably my 3rd power supply, I have RMA'd twice before. Over the course of having my computer I've almost RMA'd every part to solve my problem. I don't believe I've RMA'd my processor, could that be causing all my problems? Or am I just really unlucky with power supplies?

Side note: I ran a Windows memory test and when running the advanced test with both sticks in it said I had an error, but no error message appeared when individually tested.

Part list:
Motherboard: Gigabyte 990fx-ud3
Cpu: Amd fx-9590
Cpu cooler: Corsair h110iGT
Gpu: gtx 1080
Ram: 2x8gb corsair vengeance 1600mhz
Psu: Corsair ax760
Storage: Sandisk 250gb ssd, WD black 1tb 7200rpm
 
Solution
This came up through some searching: https://social.technet.microsoft.com/Forums/windows/en-US/308cbcb3-46ce-4f74-85f9-d87ce4cef0d6/kernel-power-event-id-41-task-category-63-spontaneous-improper-shutdowns-and-reboots?forum=w7itproperf
Check the second post. But main points:

1. Overclocking can cause problems.
2. RAM.
3. PSU.
4. Overheating.
5. Use system defaults.

With the FX-9xxx series being (heavily?) overclocked as they are, I would wonder if it's a factor. I've seen mentions of underclocking to stabilise some systems on the forums here.

With RAM, have you used MemTest? It's perhaps 'better' than the Windows one and doesn't need the OS to run. As I recall, it's recommended to allow for at least 8 passes for thorough checking. Be...
No idea.

First thing which stands out is the CPU. Given their general reputation as being unstable it makes me wonder if it is the contributing factor to the shut downs you're experiencing. Further checking shows it may depend on your motherboard. Early revisions of that model don't support that CPU according to Gigabyte's website.

Have you checked Event Viewer to see if there are any mentions of unexpected shut downs? It may provide a clue.

As for memory test... how were the individual sticks tested? Because the first explanation to come to mind is one of the slots has a problem, but you tested them individually in an error-free slot.
 

whiteguytyrone

Prominent
Sep 13, 2017
2
0
510


Just checked my motherboard, it is revision 4.0, so it is compatible.

Event view lists multiple accounts of Critical Level events, source being Kernel-Power (rebooting without cleanly shutting down), Event ID 41 and Task Category 63.

I tested the memory sticks in their original slots, however I do know the first slot doesn't recognize when any ram is in it, so I use the next pair of slots. I only let the advanced tests get to around 21% since they take so long after that. I'm not sure if I'd be able to run a full advanced test without my computer crashing. However, they both passed basic tests, individually and together.

I'd also like to note, most of the time when my computer does crash I will have to unplug it for ~10 seconds and plug it back in for my computer to start up. Even then it has crashed multiple times during startup. One problem I had last night was it would almost always crash after attempting to do a disk check or automatic repair. Both of my drives are fine and I didn't lose any data.
 
This came up through some searching: https://social.technet.microsoft.com/Forums/windows/en-US/308cbcb3-46ce-4f74-85f9-d87ce4cef0d6/kernel-power-event-id-41-task-category-63-spontaneous-improper-shutdowns-and-reboots?forum=w7itproperf
Check the second post. But main points:

1. Overclocking can cause problems.
2. RAM.
3. PSU.
4. Overheating.
5. Use system defaults.

With the FX-9xxx series being (heavily?) overclocked as they are, I would wonder if it's a factor. I've seen mentions of underclocking to stabilise some systems on the forums here.

With RAM, have you used MemTest? It's perhaps 'better' than the Windows one and doesn't need the OS to run. As I recall, it's recommended to allow for at least 8 passes for thorough checking. Be aware that such tests really check the RAM and slot at the same time rather than just the slot as such. But your comment about one of the RAM slots does make me wonder if the motherboard itself is an issue.

I would double check the CPU temperatures on full load just to definitely eliminate overheating as an issue.

At what point do crashes occur during start up? After the BIOS screen? During Windows boot up? If at the BIOS stages, then I would assume there's a hardware fault. If at Windows boot up (and given your comment about the drives), I'm thinking either a Windows installation issue (clean install perhaps?) or a faulty drive. You could try running a command prompt and use sfc/ scannow to see if it finds any issues.

That's pretty much all I can think of.
 
Solution

Neur0nauT

Admirable
If it is hanging or crashing when posting....i.e. pausing after the boot screen, Then you might have a disk issue. Check your SATA cables. In event viewer....were there any other warnings or faults before the critical power shut down message?

If you have RMA'ed two other power supplies....you might want to consider that your electrical supply to the PSU is a probable cause? If your home supply is experiencing "brownouts" it might be having a detrimental effect on the PSUs. You could consider getting a small UPS to use between your power outlet and PSU, which would narrow this down.