[SOLVED] Why am I encountering strange and random power loss problems ?

TTAOFJ

Commendable
Mar 16, 2019
7
0
1,510
0
Hello everyone, hope you have a wonderful day.
I'll list my build first:

CPU: Ryzen 5 2600
Memory: Patriot viper elite 2x8 gb 2666MHz
Mobo: Gigabyte B450 Aorus elite
GPU: Gigabyte GTX 1060 6gb
PSU: Corsair RM650x

Everything is back to stock right now(no OC)... since I encountered these problems I reverted to defaults.

My issue is really hard to reproduce because it's not happening that often. Once or twice a week I'm greeted with a sudden power loss 5 minutes after I'm into Windows. It's strange because it happens no matter of what I'm doing. I could be playing a game or just idle-ing on desktop, this problem can occur either way.

Tried stress testing after starting my pc, but,like it always happens , everything went smoothly and I didn't have any problems. I ran stress tests like prime95 combined with a gpu intensive workload even, but I couldn't reproduce the issue. Tests ran for ~4 hours without any hiccup.

Temps are fine. Max temp on my CPU was 58C in prime and GPU maxes at ~70C in valley.

When the problem does appear there are 2 scenarios:

  1. Sudden power loss with everything remaining completely dead... no restarting, no nothing. It's like pulling your power cable out and leaving it like that.
  2. Sudden power loss and after 10-15 seconds the computer starts on its own again... this happens rarely though. The first scenario occurs much more often.
I'm suspecting that either the PSU or Motherboard are at fault here, but I'm leaning towards the mobo. Can't really see why the PSU would do this when it works perfectly during max load.

What do you guys think about this? It's really frustrating for me as the problem happens so rarely and I can't reproduce it.

Any help will be appreciated 🙂.
 
Last edited:

Ralston18

Titan
Moderator
Look in Reliability History and Event Viewer for error codes, warnings, and informational events.

If you see numerous and varying errors that, to me anyway, make the PSU a primary suspect. Power problems are tricky.

Use Task Manager and Resource Monitor to observe system performance. (Use both but only one at a time.)

Boot up, open the applicable window, and watch what, if anything, changes leading up to that 5 minute mark where the power loss occurs.

Look in Task Scheduler for some app or utility being launched close to or at the 5 minute mark.

How old is that Corsair 650 watt PSU? Heavy use for gaming, video editing, or mining?

It could simply be nearing its designed EOL (End of Life) and after 5 minute or so some component heats up enough to just "quit".

Do you have a multi-meter and know how to use it? Or have a family member or friend who does?

https://www.lifewire.com/how-to-manually-test-a-power-supply-with-a-multimeter-2626158

Not a full test as the PSU is not under full load. However, any voltages out of tolerance make the PSU suspect.

And the tests you ran may have been "stress" from the perspective of the component(s) being tested.

Not necessarily the PSU itself.....
 
Reactions: TTAOFJ

Ralston18

Titan
Moderator
Look in Reliability History and Event Viewer for error codes, warnings, and informational events.

If you see numerous and varying errors that, to me anyway, make the PSU a primary suspect. Power problems are tricky.

Use Task Manager and Resource Monitor to observe system performance. (Use both but only one at a time.)

Boot up, open the applicable window, and watch what, if anything, changes leading up to that 5 minute mark where the power loss occurs.

Look in Task Scheduler for some app or utility being launched close to or at the 5 minute mark.

How old is that Corsair 650 watt PSU? Heavy use for gaming, video editing, or mining?

It could simply be nearing its designed EOL (End of Life) and after 5 minute or so some component heats up enough to just "quit".

Do you have a multi-meter and know how to use it? Or have a family member or friend who does?

https://www.lifewire.com/how-to-manually-test-a-power-supply-with-a-multimeter-2626158

Not a full test as the PSU is not under full load. However, any voltages out of tolerance make the PSU suspect.

And the tests you ran may have been "stress" from the perspective of the component(s) being tested.

Not necessarily the PSU itself.....
 
Reactions: TTAOFJ

TTAOFJ

Commendable
Mar 16, 2019
7
0
1,510
0
Just tested the PSU with a multimeter and everything is well in check.
These are the numbers:

+12V - shows 12.00V
+5V - shows 5.01V
+3.3V - shows 3.30V

Numbers are actually bang on, but as you said, without any load, those figures are not that relevant.

PSU is only 1 year and 10 months old... I also have a 7 year warranty on it so it shouldn't be anywhere near its EOL.
Agreed that stress testing is not the best approach, but that's the only way I know to make the computer draw the most power. The really strange thing is that it does the power off thing even if nothing happens... I'm guessing there shouldn't be any noticeable load when the computer is idle-ing right? Every time I stressed tested gpu/cpu or both for hours on end nothing happened. The system had a power-off either when I was idle-ing/browsing or when I was playing a non-demanding game like Age Of Empires.

I'll follow your advice with the task scheduler and event viewer this evening and report back

I'll also try starting a stress test right after turning on my computer... I never did that. Always ran them after the pc had 1 power off already.

EDIT: Checked both reliability history and event viewer and there are no errors leading up to the sudden power loss:



There is only the kernel power event and the error after restarting my computer prompting that there was an unexpected shutdown.

I don't really have many tasks scheduled... most of them are triggered at log on and are only update checks(google, geforce experience, adobe acrobat).

I also just did a quick clean, there was a bit of dust, but nothing major. What I noticed was a small greyish stain on the motherboard near the atx 24pin connector. Looked like thermal paste to me... I used Arctic MX-4 and as far as I know that's not electrically conductive.

Currently doing more stress testing after the clean... it didn't shut down on me and I'm already ~40 mins in. This is also the first time I started my PC today and if it ran for so long there's a really high chance that it won't shut down anymore. I'll probably shut it down manually and test more scenarios.

One more question: Can plugged peripherals cause a problem like this? I have a bunch of stuff plugged in... DACs, AMPs, mic, cam, controller. Right now I have only the essentials plugged in. My thinking is that there may be too much power draw on the +5V rail... though when the problem happened most of my peripherals were switched off sooo... I don't really know.
 
Last edited:

TTAOFJ

Commendable
Mar 16, 2019
7
0
1,510
0
Just wanted to make a small update on my situation.
After I cleaned everything the problem just simply vanished. 1 week passed since and no sudden power loss.

I've tried to reproduce every scenario that lead to the problem, but... it just runs smoothly now. I'm a bit in shock because there was a minimal amount of dust there, as I usually clean it once every 2-3 months. The only strange thing was that small amount of thermal paste (I'm not entirely sure if it was thermal paste though) near the 24pin ATX connector. I re-applied thermal paste on my CPU more than 6 months ago, so I don't really understand how the problem suddenly appeared after a couple of months. Strange stuff, eh.

Well, I'll mark this as solved in a few days... I just need to be completely sure that everything is working fine indeed.

Thanks for the help @Ralston18 !
 

Ralston18

Titan
Moderator
You are welcome.

My thought is that while you were cleaning etc. you may have simply re-seated something that was just a bit loose.

Or bumped it enough to establish a more stable/consistent connection.

If the problem returns then power down, unplug, and again open the case.

Verify by sight and feel that all cards, cable connections, RAM, and jumpers are fully and firmly in place.

They do tend to creep loose overtime due to heat expansion/contraction.
 
What I noticed was a small greyish stain on the motherboard near the atx 24pin connector. Looked like thermal paste to me...
This lead me to think that one of the capacitors is busted. Ref this picture <link> - You have four capacitors close to the ATX power connector. If one of them shorts, that probably explain why the PSU cut power - and it also means your PSU protection also works.

My guess is that the mobo is busted.

Just to add: If you've run the computer for a week, that may indicate that the capacitor damage are "blown clean" by several shortcuts.
 

TTAOFJ

Commendable
Mar 16, 2019
7
0
1,510
0
This lead me to think that one of the capacitors is busted. Ref this picture <link> - You have four capacitors close to the ATX power connector. If one of them shorts, that probably explain why the PSU cut power - and it also means your PSU protection also works.

My guess is that the mobo is busted.

Just to add: If you've run the computer for a week, that may indicate that the capacitor damage are "blown clean" by several shortcuts.
I've just checked the capacitors near the 24pin ATX connector and they all look healthy... no sign of leakage here:



That orange square is where that stain was. I also checked for busted capacitors everywhere on the board, but I can't seem to find any.
Everything still runs well for now... I didn't encounter any problems since I cleaned it.
 

ASK THE COMMUNITY

TRENDING THREADS