System switches off then goes into boot cycle

[CarterC4]

I have an old DP67GB motherboard, Intel I7-2600K, Corsair H80 AIO cooler, RTX 2060 GPU, MWE Gold 750W PSU, 12GB 1333MHZ RAM, 1 x 500GB SSD (boot drive), 5 x (misc) SATA drives, 5 x 120mm case fans.

System runs fine under no load, but when under load it runs for a while then switches off and goes into a boot cycle.

I've tried unplugging power from the PSU for about 5min and then rebooting it, but it remains in the boot cycle. I've tried breadboarding the mobo, but it still remains in the boot cycle.
Leaving the power unplugged overnight and then booting it up, it runs normally under no load. When under load it switches off and goes into a boot cycle again

HWMonitor temps for DP67BG seem normal (~36c) except for VREG and CPU which sometimes show max as high as ~103C and ~122C respectively. The temps for I7-2600K max out at ~28C under no load and ~60C under load for all cores.

Anyone have any ideas of what the issue might be?

 

[Trypt]

Vreg temps usually hit pretty high, 90C seems to be acceptable, but could definitely be your problem if they hit 110C. Try to log your temps and see what they were right at the crash.

 

[EarthDog]

Vreg temps usually hit pretty high, 90C seems to be acceptable, but could definitely be your problem if they hit 110C. Try to log your temps and see what they were right at the crash.

This.

Id imagine this is a temperature issue or a power supply issue since it happens under load. What 'load' are you putting it under, specifically? A CPU stress test or a game type test? If the latter, I'd lean more towards it being the PSU.

EDIT: We actually reviewed that PSU... it was MEH and said this...........

Performance
The ripple is very good, the voltage regulation is more than good enough, and the efficiency is as good as claimed. Those figures are likely to degrade as the off-brand capacitors age, however. I would also like to see better efficiency in the 50-100 W region where most modern PCs idle at, but that might not be possible without greatly increasing the complexity of the control circuit.

Design & Build Quality

Off-brand capacitors, tiny bypass relay, and evidence of cost-cutting on the inductors and transformers doesn’t bode well for long-term reliability. The assembly also shows some obvious defects. The GPU power cables putting twice as much load on the PSU side connectors as recommended by the PCIe specification isn’t exactly a good thing. The cost-cutting on the 5 V rail is asking for problems on PCs with a lot of SSDs and/or USB devices.

In summary, the Cooler Master MWE Gold 750 delivers on efficiency, but the less than impressive component selection doesn’t instill much confidence in long-term durability. It definitely would not be my pick for 24/7 use under heavy load.

How long have you had this unit? Sounds like this prediction from a 2019 review is coming true...

 

[Johan45]

Do you have any airflow in the case? It's easy to check and see if it's the VRM by attaching a fan to blow directly onto that area.

 

[CarterC4]

This.

Id imagine this is a temperature issue or a power supply issue since it happens under load. What 'load' are you putting it under, specifically? A CPU stress test or a game type test? If the latter, I'd lean more towards it being the PSU.

EDIT: We actually reviewed that PSU... it was MEH and said this...........



How long have you had this unit? Sounds like this prediction from a 2019 review is coming true...

I've tested it now using Prime 95, running torture test on all 8 threads for about 2.5 hours - see attached screenshot below. No crashes during the test.
However, moderate gaming for about ~30 minutes and the machine resets itself and goes into boot cycle.

I've had this PSU for about a year now, first time I'm experiencing this kind of problem. I'm beginning to think that your suspicions about this PSU may be correct. I'll need to test my rig with a spare PSU over the weekend and see if it does the same thing.

 

[CarterC4]

Do you have any airflow in the case? It's easy to check and see if it's the VRM by attaching a fan to blow directly onto that area.

I've got 5 x 120mm case fans , 3 IN and two OUT. Airflow in the case seems to be sufficient, I think.
I'd love to eliminate the VRM as the cause of the problem but have no idea of what I'd be looking for, or where to find it on the board. Help please?

 

[Johan45]

Doesn't sound like the VRM, P95 should have tripped that.
You can try running P95 and Unigine Heaven benchmark at the same time, this will load the PSU and would be a good indicator if it causes it to crash.