Trying To Diagnose A Problem..

jArgHA

Board: DFI Infinity P965-S
CPU: Intel e6550
Fan: Intel Stock Fan
RAM: OCZ 2GB (2x1GB, don't have exact model off hand)
PSU: OCZGXS600 (600w)
OS: Ubuntu (10.4 Lucid Lynx)

Since about 2 days ago an issue started with my system whereby it will only run for a short period of time before it just dies and begins rebooting itself (as if someone flicked the power supply switch off/on).

I have been trying to identify the problem which appears to be hardware related, but am not having much joy as of yet. The auto-reboot is *not* only happening during CPU intensive tasks, it repeatedly happens whilst I am just viewing the BIOS settings also.

I disconnected the single SATA HDD & the video card, booted up (obviously I had no video output for this test) and noted that after a minute or two the issue still occurs - that ruled the HDD and video card out of the equation I thought.

I then tried removing 1 stick of RAM, booted, same reboot issue kicks in after a minute or two. The same issue occurred with only the other stick of RAM plugged in.

I then tried running memtest86 which is bundled on the Ubuntu Live CD against each stick of RAM (with just one plugged in at a time), but after a couple of minutes of the tests running, the system reboots itself - so I am not yet certain whether the issue here could be RAM related.

I then turned my attention to the PSU, and tested the voltage rails as per the OCZ guides here:
http://www.ocztechnologyforum.com/forum/showthread.php?34941-Guide-to-testing-problematic-power-supplies.
http://www.ocztechnologyforum.com/forum/showthread.php?7069-Measuring-the-rails-on-OCZ-PSU-s.

The 12V rail read approx 12.34V, the 5V rail read approx 5.05V and the 3.3V rail read 3.43V - all these seemed OK to me - or are they?

Finally I thought that maybe the CPU temperature was getting too high, which was causing a BIOS setting to kick in and automatically shut down the system. Therefore I tried going into the 'PC Health Status' section of the BIOS and looking at the CPU temperature at the point when the system shuts itself down - the last 2 times I did this the temperature was a mere 30 degrees celcius when suddenly the automatic reboot kicked in.

Would anyone have any suggestions as to what could be the issue based on the above - I am not conclusive about the RAM but have a feeling the issue lies elsewhere. Could this kind of issue occur if there are fluctuations in the mains power supply, if so is there any easy way to check for such fluctuations?

Thanks in advance for any tips / advice!

Edit: PS Please note neither the CPU / RAM are *not* overclocked, although they have been in the past (more than a year ago).

Solitaire

Open the rig up, disconnect the frontpanel connections then use a screwdriver or spare jumper across the frontpanel header's Power pins to turn the machine on. See if the issue reappears.

One reliability isue no-one ever thinks about until it nails them - one vital component which loves to fail is the on/off switch on the case, causing Magical Rebooting Hell

jArgHA

Thanks for the tip Solitaire, going to try that now. There is both a power and reset button built onto the motherboard so it will be easy to boot without the need for jumping..

jArgHA

I tried disconnecting all of the front panel connectors from the motherboard, and the issue continues to occur. I thought for a moment you had hit the nail on the head as I was able to use Ubuntu for a little longer than I expected, but the reboot kicked in about 4 minutes after boot..

Solitaire

Your next step would be to get a second PSU or a loan of a mate's PSU and test with that. The SXS600 is a right ****e for reliability issues tbh and a fault on the 5VSB line is the most likely culprit Other than that given you're still getting this in the BIOS the next most likely culprit is the mobo, but that's not so easy to swapsies so you'd have to test everything but the mobo if you want to avoid reinstalling various CPUs half a dozen times

jArgHA

Cheers for the PSU tip Solitaire. I'm going to try to get hold of another PSU for testing to see if it resolves the issue, if not, then the dreaded motherboard change might be required...