A.R.S.E.

A problem has just started up on my desktop. I can start it up and runs fine, but then after 10min or so, it just turns off. Not a restart, just like someone pulled the cord.

Any suggestions on might be the problem? Have checked core temps and all are within range so not a thermal issue.
Is this possibly a PSu on the way out?

If there is no Blue Screen of Death or anything in the Event Logs to indicate a reason, then it could very well be the PSU.

Open it up and give it a good clean, reseat everything and try again. If the problem persists, a new PSU might be needed.

I've had that a couple of times over the years. It's been from overheating and also PSU on it's way out.

I have an old PC that was exhibiting all sorts of random behaviour. It turned out that the BIOS backup battery was dead. If your machine is old, it might be worth checking.

Alrighty, finally got round to replacing the psu, but still have the exact same problem.
It's had a new motherboard and cpu about 6 months ago as well, so shouldn't be that or bios battery.

What would be the next thing to look at?

Sounds like an overheat power down. Has the CPU fan stopped working by any chance?

CPU cores are all within good temp (40-45deg celsius) ranges while it is running, and all fans seem to be spinning still, so dont think it is a thermal issue.

Could the old graphics card or harddrive be causing issues?

I wouldn't put it down to being a problem with a GPU or HDD. Both components could cause systems to stop but not in the way your stating, on windows you would get a BSOD if either of them components were failing or things would just freeze. It wouldn't make the computer lose power all together.

Could be that the board isn't seated correctly, board is dying, you replaced the PSU so I would assume the new one is ok. Cabling, if you got a new PSU and it was modular, did you use the new cables? Same apply to power leads too.

I've seen quite a few people have weird power issues because the board wasn't seated correctly. Also, the board might be giving false readings too (on CPU temp), when it does shut down unexpectedly, is the heatsync feeling a lot hotter then the expected 45c ?

As one small and easy variable to eliminate, I'd try a different power board and power cable for the system as well. It's probably not the issue, but you never know. Not hard to remove it as a possibility.

If you built the PC yourself, did you put in all the motherboard standoffs (and are they in the correct positions?)?

Alrighty, pulled the graphics card and running the monitor off the motherboard display output.

So far, so good! Graphics card was too hot to touch after intial 15mins and crash, so guessing the fan in that had died, was overheating, and shutting the PC down.

Whats the best bargain graphics card out there at the moment to replace the above one that has finally give up after 7 years hard slog.
Looking at around $250 as I dont really do much gaming these days to justify going large!

thornz wrote:Alrighty, pulled the graphics card and running the monitor off the motherboard display output.

So far, so good! Graphics card was too hot to touch after intial 15mins and crash, so guessing the fan in that had died, was overheating, and shutting the PC down.

Over the years I've had 2 cards with hardened thermal paste which renders the card pretty much useless. Maybe redo the paste and see if that helps before buying a new card if your cash strapped.

Sent from my Nexus 5