Question Computer Randomly Shuts down

[TheDudePlayingaDude]

Hello, so, i 'have' this computer that started to randomly shutting down (no BSOD). It started while on demanding operations,
such as streaming plus gaming with dual monitors, but it also happens while idle, and i'm needing some help trying to find out why.

System:
Windows 10 Pro 64x

Specs:
Mobo: Gigabyte Z87X-D3H-CF
CPU: Intel core i5-4670 @3.40GHz
GPU: NVIDIA GeForce GTX 760
RAM: 2x 4gb Corsair Vengeance DDR3
PSU: 700W TR2 Thermaltake


What has been done:

CPU: hours of 100% stress test and never shutting down during those. First test high temperatura were as high as 100ºC, but after a good clean up plus
changing thermal paste, it doesnt go higher than 75ºC, with idle temperature at 35º, ranging up to 60 on regular semidemanding such as my gaming.
It still can shut down while idle @35ºC, but it has never done so while on those hours long stress test.

GPU: Has been stressed for hours without ever shutting down during those, and has been taken off the equation with the monitor connect to the mobo.
Problem still persisted.

RAM: Windows memory diagnostic says it's fine. 8 hours long MemTest says it's fine. Have already switched to the other dual channel position and
tested each at a time. Problem always persisted.

PSU: Have tried to switch the PSU to another one that is working smoothly (600W) and the problem persisted, one could argue that the 600W wouldn't be
enough, but the thing is that the other computer is WAY more wattage demanding than this one (450W vs 550W calculated demands), and is ALWAYS on,
never having an issue. Problem persisted.

System: Has been clean formatted, all drivers thoroughly updated, fast boot disabled. Problem persisted.


What i didn't really test:

Changing Mobo, as i don't have a spare one.

Changing CPU itself or testing it in another mobo, for the same reason.

HDD, since i saw no reason for it, and it seems to be working perfectly, plus no issues on basic diagnostic tools, and it has shutdown once (right after a shut down - twice in a row) before accessing it.

Updated BIOs, as i have never done that and i know there can be permanent damage.



So, any thoughts, guys?
Any help is really appreciated.

Thanks!

 

[clarkjd]

Freezes tend to be memory related, while shut-downs and re-boots tend to be power related. That being said, Thermaltake TR2 series are not what is considered a quality PSU series. What brand, make and model was the other PSU that you tried?

 

[TheDudePlayingaDude]

Freezes tend to be memory related, while shut-downs and re-boots tend to be power related. That being said, Thermaltake TR2 series are not what is considered a quality PSU series. What brand, make and model was the other PSU that you tried?

OCZ600MXSP, which is the one i'm using in this computer, which i use for heavy gaming and has never failed

Specs:
mobo: gigabyte 970A-UD3P
CPU: AMD FX-8350
GPU: Radeon R9 280X
RAM: 2x 8GB Cosair Vangeance
PSU: OCZ600MXSP
SSD + 2xHDD, quadMonitor...

Both computers are considerably old, but this one is more power demanding. I initially would've guessed PSU or heating CPU too, but given the tests, sounds like a stretch.
Maybe some permanent damage to CPU from previous almost 100ºC temperatures? idk...

 

[clarkjd]

OCZ600MXSP, which is the one i'm using in this computer, which i use for heavy gaming and has never failed

Specs:
mobo: gigabyte 970A-UD3P
CPU: AMD FX-8350
GPU: Radeon R9 280X
RAM: 2x 8GB Cosair Vangeance
PSU: OCZ600MXSP
SSD + 2xHDD, quadMonitor...

Both computers are considerably old, but this one is more power demanding. I initially would've guessed PSU or heating CPU too, but given the tests, sounds like a stretch.
Maybe some permanent damage to CPU from previous almost 100ºC temperatures? idk...

Once a CPU is damaged, it won't randomly work. There are built-in protocols that prevent CPU damage by overheating.
What does "considerably old" mean, exactly? Is either PSU still under warranty?

Another possibility: a bad (intermittant) connection from the PSU to the motherboard, either the 24 pin main power connector or the EPS connector for the CPU.

 

[TheDudePlayingaDude]

Once a CPU is damaged, it won't randomly work. There are built-in protocols that prevent CPU damage by overheating.
What does "considerably old" mean, exactly? Is either PSU still under warranty?

Another possibility: a bad (intermittant) connection from the PSU to the motherboard, either the 24 pin main power connector or the EPS connector for the CPU.

I think it's safe to say that both have been in use for more than 5 years, can't say exactly though.

In case of the intermittent connection from PSU to the motherboard, could that problem be caused from the motherboard connector?

If i didn't test the good PSU i would be 99% sure it's the PSU by now, but it showing exactly the same symptoms is really preventing me to be so sure, thus being
afraid to buy a new one and nothing change. Guess i'll try to test the good PSU on it again with the GPU disconnected, so there DEFINITELY is enough power.
But if that doesn't do it, it just yells to me that the PSU isn't the problem.

 

[clarkjd]

I think it's safe to say that both have been in use for more than 5 years, can't say exactly though.

In case of the intermittent connection from PSU to the motherboard, could that problem be caused from the motherboard connector?

If i didn't test the good PSU i would be 99% sure it's the PSU by now, but it showing exactly the same symptoms is really preventing me to be so sure, thus being
afraid to buy a new one and nothing change. Guess i'll try to test the good PSU on it again with the GPU disconnected, so there DEFINITELY is enough power.
But if that doesn't do it, it just yells to me that the PSU isn't the problem.

It could be that the 24-pin connector on the motherboard is defective and therefore causing an intermittant connection.

 

[TheDudePlayingaDude]

Just reporting to add some insight for further people with the issue and the backers. The motherboard was indeed the issue, but not related to the pins. Some part of it was overheating extremely (even though it reported fine temperatures) to the point it even got some liquid leaking, near its thermal dissipator.

Thread done.

Thanks.