[SOLVED] How to fix seemingly random and inconsistent 'crashes' (no blue screen, no errors), after which power button is unresponsive?

Sep 11, 2020
3
0
10
0
Last weekend, I was in the middle of a Skype call when my computer (Win10 Pro) just suddenly powered off (after being on for around 1.5 hours). No blue screen error nor a beep code from the motherboard: the fans just stopped spinning suddenly and the screen turned off.

Something similar had occurred the previously (2 weeks ago) with 3 or 4 crashes seemingly at random, although these were all quite soon after logging into Windows. I thought it might be a software issue with some sort of rogue process running on startup, and, after I changed some power settings (including turning off Fast Startup), the issue seemed to go away.

However, this new crash led to the power button being unresponsive - pressing it had no effect at all (which hadn't been the case previously - I had been able to restart the computer again fine).

Only after unpowering my PSU (flicking the rear I/O switch) and turning it on again did the power button actually do anything. I got into BIOS, loaded its fail-safe defaults and rebooted the PC. It crashed (in the same manner as before) during Windows load (with Win10 logo still onscreen and loading circle underneath). Again, the power button ceased to work after this crash.

Potentially notably, two LEDs (one green, one yellow) had illuminated on my motherboard (GA-EX58-UD5 (rev. 1.0)) next to the RAM slots after the crash. According to the manual (manual URL) these indicate 'Memory (DDR Voltage)'. So, after unpowering, I swapped out my two 4GB RAM sticks (both and individually) (Qumox 8GB (4GBx2) 1600 DDR3) and tried booting (having to switch the PSU off and then on again each time), every time unsuccessfully. Interestingly, every time I turned the PSU on again and the motherboard was powered, the same two LEDs flashed briefly. According to page 39 of the manual this indicates DDR loading and North Bridge loading; however, I don't know what this means in practice, if I've read the manual correctly or even whether they normally illuminated in the past before these issues started occurring (sorry :-( ).

At this point, I decided to turn off completely again and leave the computer for a while since I didn't have time at that moment. I came back to it around 5 hours later with a friend and he suggested thoroughly de-dusting the entire thing. We did so, taking out most components in the process, including the GPU (NVIDIA GeForce GTX 680), also loosening some wires which appeared quite tightly squeezed thanks to some aggressive zipties from the person who had built the computer. After all this, we tried booting and I managed to get all the way into Windows and seemingly fully operational. I was then using the computer normally, thinking we'd fixed the issue, but, just in case, decided to setup hardware logging with MSI afterburner.

Sure enough, after around 3.5-4 hours, it randomly crashed again - I was in the middle of watching a YouTube video this time. This time, after crashing, the computer tried to power up 2 or 3 times of its own accord with the fans spinning up and the motherboard lighting up. However, on no occasion did it post and after these attempts it gave up - power button now once again unresponsive.

I left it overnight (giving it over 12 hours of 'rest') and in the morning it booted again - only to crash after 1hr 15 mins (I'd timed it this time). Only 2 chrome tabs were open at the time. Hardware logging had revealed nothing. Logs were being written until literally the second before the crash with all values (RAM usage, RAM voltage, GPU usage, CPU usage, CPU temp, GPU temp, CPU clock speed, Power usage, etc.) normal and nothing out of the ordinary. Windows event viewer and control panel reliability history had nothing for me either in terms of any potential software errors. (CPU is Intel Core i7 920 @ 2.67 GHz (not overclocked))

Over the next couple of weekdays, I only used it in the evenings. Some evenings, I was able to use it seamlessly for 2.5+ hours playing games, watching videos, using Discord, etc. Other evenings, I got up to around 30 minutes of use before a crash. I now switch it off at the PSU every night after a crash or just a regular shutdown if I had been lucky that day.

Today, after working perfectly yesterday evening for 3+ hours, it crashed again, a while after logging on (~10 minutes after first power on) and the power button was, as usual now, unresponsive. The pattern appears to be work, crash, work, crash...

My friend suggested I should make a post about it and then hope, and here I am.

NB: I don't have a PSU of sufficient power on hand to replace the current one (Corsair TX 650W) if it is a PSU issue. Although the fact that the computer tries to restart itself sometimes, leads my friend to believe that this isn't the issue.
 

Ralston18

Titan
Moderator
Symptoms suggest a PSU issue.

How old is the Corsair TX 650 Watt PSU? Was it new, reconditioned, used when first installed?

Possibly heat related - warms up, fails, and then works when once again cooled down.

Trying to restart and failing could simply be some component within the PSU has gotten too hot affecting PSU performance.

Remember that the PSU provides varying voltages and in only takes one of those voltages to go out of spec and crash the system.

You could test the PSU:

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

Not a full test per se as the PSU is not under load. However, any voltages outside of spec would indicate failing PSU.
 

Ralston18

Titan
Moderator
Symptoms suggest a PSU issue.

How old is the Corsair TX 650 Watt PSU? Was it new, reconditioned, used when first installed?

Possibly heat related - warms up, fails, and then works when once again cooled down.

Trying to restart and failing could simply be some component within the PSU has gotten too hot affecting PSU performance.

Remember that the PSU provides varying voltages and in only takes one of those voltages to go out of spec and crash the system.

You could test the PSU:

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

Not a full test per se as the PSU is not under load. However, any voltages outside of spec would indicate failing PSU.
 
Sep 11, 2020
3
0
10
0
Symptoms suggest a PSU issue.

How old is the Corsair TX 650 Watt PSU? Was it new, reconditioned, used when first installed?

Possibly heat related - warms up, fails, and then works when once again cooled down.

Trying to restart and failing could simply be some component within the PSU has gotten too hot affecting PSU performance.

Remember that the PSU provides varying voltages and in only takes one of those voltages to go out of spec and crash the system.

You could test the PSU:

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

Not a full test per se as the PSU is not under load. However, any voltages outside of spec would indicate failing PSU.
Thanks for the quick response!
I got the computer from a friend and have been using it myself for around 2 years with this power supply. He'd previously been using it for around 2-3 years before I got it I believe.
I'll test the PSU as you suggest tomorrow and get back to you on this thread.
What could the reason be the alternating day cycle though - as in, after having over 20 hours to cool down in both cases, why does it work one evening and not the other?
 

Ralston18

Titan
Moderator
Could be multiple developing problems. PSU and/or some other system component.

Some combination of events and circumstances. Room temperature, what games you play, what the computer is doing in the background.

Intermittent problems are the trickiest. Alternating cycle pattern may be coincidental.

There could be pattern but it would take a very controlled (i.e., only one change at a time) testing effort to find a pattern that identifies a specific source reason for the crashes. I think a much longer time frame would be needed to establish that cycle as a true symptom.

You might take another look at the logs - look for something that precedes or corresponds with the crashes.
 
Sep 11, 2020
3
0
10
0
Just an update for anyone in the future:
I bit the bullet and bought a new PSU (a TX650M Corsair).
I will post here if the computer crashes again with this new PSU installed but otherwise, assume that this fixed the issue!
 

ASK THE COMMUNITY

TRENDING THREADS