Run MemTest86 - run it a lot.
Remove the graphics card and use the on-board video instead.

I would try to determine if it's truly hardware or not, first.

Get a copy of memtest86 and let it test your memory. Although, the fact that it reboots while idling would lead me to believe it's something else... like cooling or dying power supply.

Use a stress tester like Prime95, OCCT, or IntelBurnTest to stress the CPU out. If it's cooling or PSU, one of these utilities will cause the system to lock up in no time at all. OCCT, in particular, keeps track of your system's voltages during it's tests. So if you lock up, with any luck, you'll be able to see if your voltages were starting to drift from what they're supposed to be.

9/10 times, this sort of thing is either bad memory, cooling, bad or failing PSU, or hardware or software incompatibility. Since you've used this hardware in the past with no issues. My guess is one of the first three I mentioned.

It must be cooling. Just try to lower voltage and speed of the CPU (underclock) a bit and see if the problem is still there.

or it is very very dusty in side?

Thanks for these suggestions.

So far I have tried prime95. It seems as though once the core #0 temperature reaches about 70 degrees for a while, the system reboots. The small FFTs and the in-place large FFTs tests give the quickest temperature rise and therefore the fastest reboot. Which shows that my cooling can't keep up with the heat produced by prime95, though I can't see why the system would get that hot normally - right now Core #0 is at 42 degrees. From what I have found on the web, the athlon 5400+ maximum temperature is 70 [I can't tell whether that's per processor or the average, but it seems when mine stays around 70 for a while it crashes]. Guess I need to keep track of the cpu temp in normal usage.

Will try a few other things now - I have memtest86 on a boot cd somewhere. Hoping it's not the psu as that's pretty new, but I guess it might not be up to the job.

If a stress-test causes reboots, then you definitely want to lower the clock-freq, or improve the cooling.

did you have any thing to do or apply the thermal grease on the cpu? if so what kind?

You should configure it to BSOD instead of rebooting: Hit win+pause to bring up the System Properties dialog, then click the Advanced tab and click Settings under Startup and Recovery. Now, clear the "Automatically restart" checkbox and hit OK on both dialogs to close them.

You should now get a blue screen of death(TM) with some diagnostic info instead of a reboot. If you don't get a BSOD then you may have a bad PSU or your CPU may be overheating (usually that causes a shutdown rather than reboot though).

Blow the dust out of your case and fans and make sure the fans are all working.

some thermal greases dry out like Arctic silver 5 they either have to reapplied or a or a longer lasting compound used i think some stuff called mx something or some such thing is almost as good as AS5 but, last much longer.

HunterZ wrote:

You should configure it to BSOD instead of rebooting: Hit win+pause to bring up the System Properties dialog...

Hey thanks for the shortcut!!! Makes it easier to explain "right click on my computer, then properties" sentence! 😅

cdoublejj wrote:

some thermal greases dry out like Arctic silver 5 they either have to reapplied or a or a longer lasting compound used i think some stuff called mx something or some such thing is almost as good as AS5 but, last much longer.

I personally have better results using Arctix MX-1 instead of Arctic Silver 5. I think AS5 is a bit over-hyped and not as perfect as people make it out to be.... compare it with MX-1 and see for yourself.

Arctic claim after a period of 200 hours you will notice your temps drop slightly more with AS5.... I never saw any change after months since the day i applied it 😒 I now only use Arctic MX-1 but there still could be better stuff out there?

I bought a giant 22g tube of Arctic Silver Ceramique for cheap. It cost only a couple $$ more than a 3.5g tube of AS5. It's actually one of the few good bulk options available to consumers.

The differences are negligible between the best and worst pastes. What matters most is how you apply the compound. Arctic Silver themselves have some really great tutorials on that btw.

i use arctic silver for even my little projects like p3 and celery processors AS5 is better than the cheap whit stuff. MX-1 was the stuff i was thinking if i reapply i'll use it.

Thanks for all your help and suggestions, as a result I think I've solved it now.

I spent some hours on this problem today, checking for dust, re-appling heat paste etc., and re-checking my understanding of what the temperatures ought to be within - this was a red herring I think, my fault.

The system rebooted rather than shutting down [usually anyway] and failed in OCCT with an error [not specified as far as I can see]; failed after a while in prime95; and threw up errors after 12 minutes in memtest86.

I took out the memory sticks and tested them individually. Memtes86 then prime95. No errors. Then again together. No errors [yet]. So maybe it was just the memory needed reseating; this had been one of the first things I did before posting here. Memory seems pretty loose in this motherboard, none of that sharp snapping into place like you got with the old PC100/133 sticks.

Since you have noticed that the CPU was at 70 degrees celcius at some point, you should definitely upgrade your cpu cooler. The hotter the cpu gets, the sooner it will kick the bucket. Have a constant check of your cpu temperature.

BUT:

So far I have noticed only shutdowns when the cpu gets too hot but never the reboots. If it simply reboots, it's not the mobo's auto-shutdown mechanism at work, since rebooting will aggravate the already hot cpu temperature, not reduce it.

I'm not so sure if the memory seating is the problem here which is causing the reboots. Most of my unknown reboots are due to mobo-graphics card incompatibilities or certain driver upgrades. Better to downgrade a graphics drivers to a more stable version or wait for the next driver release.

The number one cause of reboots in my case, is the FSB overclocking. My latest experience is with the Striker II Formula Motherboard. Overclocking the FSB even a little causes random reboots at no particular moment. I'm very particular about cpu cooling and cpu case air circulation.

My main system is a quad core Q9300, with peak running temp. of around 43-45 degrees celcius, and maintained at that level. At idle, it runs around 37-39 degrees. The heatsink you see (or can't 🙄 ) below is a Cooler Master Gemini II. Bought two silent 120mm fans (which are fixed on the Gemini II Heatsink), blow down on to the cpu and mobo. The ram sticks have only stock heat spreaders, each running at 1066MHz at 2.2v. The cables here are not tied properly yet.

Sounds like a ram issue to me, the power supply may be defective, the video card may be defective, some other piece of hardware may be causing a buss conflict. Have you updated the bios lately? There also may be dust bunnies behind the motherboard. The ps2 port might be shorting out (seen this before). There can be any number of reasons... tracking it down is the hard part. Also check the bios battery, sometimes they go bad and short out/arc randomly aswell.

Checked for rootkits/viruses? some virii and rootkits will reboot the machine at random (especially if it's a bot and under someone elses control), but you didn't say if you merely upgraded win2k to xp or did a clean install , so I'm just suggesting this.

my water cooling system leaked sludge on my video card it wouldn't even quit it just bugged up made some really cool/weird artifacts i sitll use i just cleaned it off with rubbing alky and put in my now air cooled phenom machine.

Random problems are the worst. My last one turned out to be a problem hard drive that had passed every test available. I replaced it on a hunch because it seemed programs were taking an extra second or so to load.

Before that I had a heat problem caused by a cpu fan that would stop for a minute and restart. Now when I troubleshoot I not only check if the fans are spinning on startup, I also check them throughout the day.

And thanks RoyBatty for the tip about ps2 ports. I’ll add this to my list of thing to check.

The problem recurred.... I got better case cooling installed, which eliminated that in a rather deafening way, and then I found that one of the memory sticks was indeed failing, so I guess this was the real problem. Replaced the RAM and not had a lockup since.