Random power cuts on desktop

umkay · 13 Dec 2020 at 21:06

Hi all,

Long time lurker here in need of some help. My desktop will cut out power completely (all monitors black out and all fans/activity seems to stop) and then power itself back on almost immediately and start posting. Sometimes it happens a few times a day and sometimes it won't happen for a week or two. I'm usually pretty adept with hardware issues but not entirely sure where to start on this one.

Notes:
- Built this PC in 2014 and upgraded the GPU in the past few months
- This restarting started happening sometime last year
- Temps look normal
- Happens in all scenarios - web browsing/gaming/computer left on overnight
- Windows reliability monitor shows the event as 'Windows was not properly shut down'

Specs as follows:

i7 4790K @ 4.0Ghz
32GB DDR3 1600 Corsair Vengeance Red RAM (4x8GB)
Gigabyte Z97X-UD5H-BK Black Edition 1150 Motherboard
MSI Nvidia GTX970 4GB
750W Corsair RM Series 80PLUS Gold Modular Power Supply

My gut is leaning towards PSU but I feel it could also be mobo or RAM? Any help is appreciated, thanks in advanced!

Blue160 · 13 Dec 2020 at 21:27

Sounds very like the psu to me. All power cut with no blue screen, it’s the obvious culprit.

The only way to confirm is to try a different psu.

mickyflinn · 13 Dec 2020 at 21:32

Intermitent faults are a pain but makes sure everything is set to stock settings.

Sounds like the psu what model is it ?

umkay · 13 Dec 2020 at 21:53

Blue160 said:
Sounds very like the psu to me. All power cut with no blue screen, it’s the obvious culprit.

The only way to confirm is to try a different psu.

Thanks Blue160, you confirm my suspicions on the PSU. I guess you're right, the only way to find out is to replace the PSU and go from there. I just hope it doesn't into a money pit ...

mickyflinn said:
Intermitent faults are a pain but makes sure everything is set to stock settings.

Sounds like the psu what model is it ?

Thanks mickyflinn. I've running everything stock and I've got the 750W Corsair RM Series 80PLUS Gold Modular Power Supply. I think it's just out of warranty too unfortunately.

postmanfw · 13 Dec 2020 at 21:56

Have you checked the dust filter on the case for PSU isn't clogged up? (if it has one)

BUDFORCE · 13 Dec 2020 at 22:14

postmanfw said:
Have you checked the dust filter on the case for PSU isn't clogged up? (if it has one)

Or the GPU for that matter.

Doh didn't read first post.

Yes probably power supply, could be a capacitor on motherboard failing. I'd try reseating stuff first.

umkay · 14 Dec 2020 at 23:16

postmanfw said:
Have you checked the dust filter on the case for PSU isn't clogged up? (if it has one)

The case is pretty clean, but I checked and nothing seems clogged.

BUDFORCE said:
Or the GPU for that matter.

Doh didn't read first post.

Yes probably power supply, could be a capacitor on motherboard failing. I'd try reseating stuff first.

Reseating RAM, CPU etc?

Also any recommendations on a replacement PSU?

BUDFORCE · 15 Dec 2020 at 06:45

BUDFORCE said:
Or the GPU for that matter.

umkay said:
The case is pretty clean, but I checked and nothing seems clogged.

Reseating RAM, CPU etc?

Also any recommendations on a replacement PSU?

Probably wouldn't bother CPU, if you do be really careful, I've seen it happen the thermal paste acts like a glue and when you pull the cooler out, it pulls the CPU also which can damage it.

RAM sure, but more importantly go round check all the cables are properly in. It's just worth a go before spending money.

Edit: that includes the main power cable that goes into the back of the PSU and the plug into the wall

Is this getting worse by the way? More frequent?

wolfie138 · 15 Dec 2020 at 08:13

meh, this sounds like the issue i had, it can drive you insane. puter would run for months no prob then randomly just power off and come back on. nothing noteable in the event log, no faults on the HDDs according to scans/tests, no noteable temp/voltage fluctuations in HWInfo. i was always convinced it was a glitchy ASUS board.

Joxeon · 15 Dec 2020 at 08:17

Have you checked temps to make sure those are ok?

BUDFORCE said:
Probably wouldn't bother CPU, if you do be really careful, I've seen it happen the thermal paste acts like a glue and when you pull the cooler out, it pulls the CPU also which can damage it.

Not really an issue with Intel as the CPU is held in place with the locking mechanism.

Quartz · 15 Dec 2020 at 08:58

I had this same problem. It wasn't the PSU. It was the CPU or motherboard.

umkay · 15 Dec 2020 at 12:52

BUDFORCE said:
Probably wouldn't bother CPU, if you do be really careful, I've seen it happen the thermal paste acts like a glue and when you pull the cooler out, it pulls the CPU also which can damage it.

RAM sure, but more importantly go round check all the cables are properly in. It's just worth a go before spending money.

Edit: that includes the main power cable that goes into the back of the PSU and the plug into the wall

Is this getting worse by the way? More frequent?

Thanks for the heads up

I'll have a look at all the connection points as you mentioned. I did strip the PC down, clean the case and put it back together last year. Could that have caused this issue?

The frequency is bizarre, sometimes once or twice a day, sometimes nothing for a week or two :S

wolfie138 said:
meh, this sounds like the issue i had, it can drive you insane. puter would run for months no prob then randomly just power off and come back on. nothing noteable in the event log, no faults on the HDDs according to scans/tests, no noteable temp/voltage fluctuations in HWInfo. i was always convinced it was a glitchy ASUS board.

Yeah it's driving my insane too tbh. All it says is that the pc was unexpectedly powered off.

Joxeon said:
Have you checked temps to make sure those are ok?

Not really an issue with Intel as the CPU is held in place with the locking mechanism.

Yeah they all look good. CPU is around 60*C, mobo is 45*C, GPU 60*C, SSD 40*C and HDDs 30*C.

Quartz said:
I had this same problem. It wasn't the PSU. It was the CPU or motherboard.

Ahh I'm so lost on which part to buy first. Since the socket 1150 is legacy now, it doesn't seem worth reinvesting into this setup if it's not a PSU issue?

Mrwong · 15 Dec 2020 at 15:44

I have the same problems sometimes it just goes into a loop.before it reaches the boot screen other times it just goes into windows works the suddenly cuts out and constantly restarts.

I about to buy a psu to see if that is the issue. If not then it's a new board chip memory.

I am still running win7 the is after that I didn't like the gui.

umkay · 15 Dec 2020 at 19:06

I just noticed my CPU temps are 96*C CPU load is about 60%. I'm using the stock intel cooler as my Corsair H80i broke. Could those temps be causing any issues?

adrichardson81 · 15 Dec 2020 at 19:19

Yeah that would definitely be a problem. From memory my 4770k didn't go over 75c torture testing at 4300.

BUDFORCE · 15 Dec 2020 at 21:34

umkay said:
I just noticed my CPU temps are 96*C CPU load is about 60%. I'm using the stock intel cooler as my Corsair H80i broke. Could those temps be causing any issues?

Yeah thats really not good.

umkay · 15 Dec 2020 at 21:57

adrichardson81 said:
Yeah that would definitely be a problem. From memory my 4770k didn't go over 75c torture testing at 4300.

BUDFORCE said:
Yeah thats really not good.

Could it just be that the stock cooler can't cope?

adrichardson81 · 16 Dec 2020 at 20:06

umkay said:
Could it just be that the stock cooler can't cope?

Pretty sure I had an aio with mine, so it could be too much for the stock cooler.

alphaomega16 · 17 Dec 2020 at 03:49

umkay said:
I just noticed my CPU temps are 96*C CPU load is about 60%. I'm using the stock intel cooler as my Corsair H80i broke. Could those temps be causing any issues?

Sweet jesus 96c, wow.

Blue160 · 17 Dec 2020 at 07:50

umkay said:
I just noticed my CPU temps are 96*C CPU load is about 60%. I'm using the stock intel cooler as my Corsair H80i broke. Could those temps be causing any issues?

Yes!

Even with the stock cooler you shouldn’t be hitting those numbers. I’d start by checking all four legs of the cooler have clicked into place properly.