Over the last 6 months or so my computer will have spouts of instantly turning off. When I boot back up, it will beep a few times then say that it was turned off due to CPU temperatures. But the problem is that I will be watching the CPU temp in a monitoring tool (or through the BIOS) and will be anywhere from 40 to 60 degrees when it shuts off, which is obviously within limits.
Motherboard: Intel D865PERL
CPU: P4 3.0ghz, 800mhz fsb, Prescott (ugh, hot!)
Power supply: Antec 430 watt truePower, which is connected to power through an UPS.
The odd part is that it comes in spurts. Everything will be fine for days, weeks or months, then out of nowhere it will start turning off. And it will continue to do it for hours or a few days, then magically be okay again. When it is in one of its tantrums, it will reboot anywhere from 1 second to 5 minutes of turning my computer on. (It will literally do it within a second or two of powering on the computer) Originally I thought it was the CPU over heating, but then saw the CPU not getting beyond its limits. I then thought a faulty CPU, Intel sent me a replacement, but problem persisted. And thats the point I am at now.
I really don't know what else it could be. Could it be the powersupply failing, or spiking, causing the CPU to overheat? could it be the chipset of the motherboard overheating? Or some aspect of the motherboard?
31computers: Have already reseated everything, and cleaned the cpu/contact point of heatsink, then reapplying thermal paste. Could the beeps still mean bad memory if it says "Computer was shut down to CPU thermal issues (CPU overheating)"? or something similar? I have reset the memory and run MemTest and neither showed anything wrong Have also turned off the automatic fan control so it never goes lower than high.
Anoobis: I couldn't find that setting, or even a way to view what this threshold is at. And since that setting isn't there I just went by what Intel support said, which was that it was set to shut down the system at 75 C
Also, I have run CPU Burn-In for awhile trying to see if the CPU would get too hot and turn off. And with it running at 100% capacity for over an hour it didn't turn off. Only to turn off about an hour later during normal use.
Hmm. Perhaps there is something wrong with the motherboard. Have you visually inspected it for any signs of damage like bulging or leaking capacitors? Has the board been subject to any brown outs, black outs or power surges?
I did look over the motherboard a few days ago. Took it completely out infact so I could get a look at the back as well. There was nothing I could spot, but really wasnt sure what I was looking for.
Just an update, I changed the powersupply with an old Antec 300watt, and its been okay since I did that. But since the problem is so random I am somewhat expecting it to come back anytime now. I hate the feeling I get when I think of that, or when I catch myself getting my hopes up =P
And btw, thank you to everyone who has replied so far
"A few beeps" could be anything. You haven't determined the maker of the BIOS, each one has different beepcode. You haven't determined how many and how long the beeps are. That's why they call it a code. Each code means something different. You are ASS-u-ming and the person you are advising will pay the price.
I kinda didn't want to bring the PSU into this until the last resort. It seems like everyone around here is so quick to suggest it's a PSU problem that it has become sort of a generic (and costly) answer. The problems you indicated could be a PSU issue but the Antec TruePower series has a very good track record and I didn't want to bring it up until other options had been suggested although the randomness of your issue is a strong argument that favors the PSU as the problem.
Two days now that Ive had this powersupply in there, and still no shut offs So I'm hopeful. And I found an Antec 500watt (wanted something I could upgrade with the next time I upgrade) for $50 online (after a mail-in rebate) so Ill hopefully get that soon and have this fixed all together