What is a good fault finding strategy using HWInfo?

Hi,

My Windows10 PC has started intermittently crashing without a BSOD. It's been fine for years. Now, I can use it under load for hours without a problem until one day it will crash. Sometimes twice. I have to manually power down each time. When I power back up, everything is fine. Most notable during these crashes is that my two displays go blank, the disk activity light still flashes and some lights on my motherboard stay lit, meaning there is still power being supplied to the board.

The Windows event viewer and Reliability tool only report a LiveKernel event (141 among others, it varies) and the computer losing power unexpectedly.

I do not overclock and nothing overheats. I have reseated the GPU, cleaned the contacts. Reseated and cleaned the GPU power connectors (each end). I have used FurMark to stress test my GPU but it doesn't really get too hot (between 60 and 70 celcius) and runs fine. I have run memory tests for hours without a single fault.

I have started logging temperatures and voltages and inspected the results after each session. I currently have one log taken during a crash. I expected to see at least one clear indicator of a possible problem, but after reviewing all of the sensor data I can find nothing that seems out of the ordinary. All of the values seem just fine.

Who can suggest a good strategy to continue my search for what is causing my crashes? Where do I go from here?

Any useful tips would be appreciated.
Thanks
 
Update:
One thing HWInfo doesn't give me is the PSU temperature (haven't checked fan speed) for my Gigabyte Aorus AP750GM PSU. As it turns out it has a fan on the bottom of it. My case has a filter located there which was completely clogged with dust. The PSU overheating and shutting itself off would explain the behaviour I was seeing. It did tend to happen after 3 to 4 hours of load.
Filter is cleaned. PSU has been vaccuumed. Now I'll wait and see if it happens again.

Question now is: is there a way in HWInfo of telling PSU temperatures, behaviour, fanspeed?
 
Back
Top