odd behaviour

so i got this packard bell all in one pc thingy dumped in front of me, the owner had upgraded to windows 8 and started getting random shutdowns.

so i started some diagnostic stuff on it checked the usuals before turning it on, checked the ram and hdd on a seperate test platform.

turned it on and all seemed fine booted into windows no problem, popped over checked the event log, filtered it out and found lots of thermal shutdowns. i stopped there.

now this is where it gets fun.

as the case was still open i checked the fan and the heatsink, all clear and working fine, checked the thermal paste and made sure they were in contact with the cpu.

slotted in a temp meter to check the cpu/gpu.

turned it on and back into windows 8, decided to run the wei for a minor stress test(hadn't been run anyway) starts and gets to running dx10 rendering then shuts down claiming a thermal issue.

checked the temps the entire time, got to 45-50 deg Celsius, no where near the shut off or thermal cap of the parts.

now as to why i'm posting it here on the windows 7 forums, i stuck another hdd in and loaded windows 7 ran the same test, at the exact same point it falls over, I was hoping it was just a driver issue with windows 8 but i'm beginning to wonder now.

i did do other checks but now i need some brain storming from awesome people as i am probably missing or overlooking something silly and basic and can't see it for looking

any thoughts or things to try would be great:)

From what I've been able to find out, everything is pointing to the thermal paste or the lever that locks down the cpu in place is loose and needs to be reseated.

So, my advice would be to check and recheck all the physical connections, even the cables (all) of that PC. It sure sounds like something is loose either with the cpu, the cpu fan and heat sink.

worth a check, i'll have a look tomorrow:)

it's possible the fan isn't spinning enough but the temps were on a meter the whole time, i needed a more accurate reading and one that would show the temps for what they were.


i'll check with that as well fan doesn't spin that fast but then it's not getting that hot in actual temp, need as many options as i can think of xd.

it's possible the gpu just burnt out already and i'm getting a false reading now after the fact.

cheers for thought's so far guys much appreciated, i'll let you know tomorrow if any of that showed any new info.

thanks all for the help, found the problem.