Main rig down - may be PSU

[vulcan500rider]

Hi All,

I just had my main rig go down, and I've done some basic looking at it, but want someone to check my logic for me.

Now, initially it just turned off spontaneously while folding/streaming video. I attempted to turn it back on, and it would boot, but not POST. the 780i has BIOS code PS, or 5d, depending on which way is supposed to be up on the damned display. No code for PS, but 5D is SuperIO. OK. So I clear CMOS and remove all peripherals except keyboard and mouse, which I put into different USB ports.

It boots! HOWEVER, if you so much as move the system, it shuts off, and it shut off without moving it slightly after starting Windows...Awesome.

This, of course, leads me to believe that it may be a short, but that seems unlikely, as the system has been running fine, and completely unmoving for a while now (IIRC, the last time I moved it was early Jan when I swapped my e5200 for a Q9550) months now. The only thing that has changed recently has been an OC of the 8800GT SLI cards.

This, in turn, leads me to guess that I may have killed the PSU, and I am currently going to remove it to test things further. I'm leaning toward that as I know I was probably pushing my luck running OC'd 8800GT SLI on a 500W PSU alongside an OC'd 9550...

Any other thoughts, guys?

Cheers!

Oh, and my full system specs are listed top right.

[vulcan500rider]

Update: Got it on my testing bench and opened everything up. I tried to recreate the problem with no success.

I tried voltage testing with a multimeter and everything looks rock solid.

I put the second GPU back in and booted. POST now says "Warning! Now system is in safe mode. Please reset CPU or Memory Frequency in CMOS Setup.

I'll clear the CMOS again, but I would be surprised if that was it; I just reset it a few minutes ago, after all.

[MarkOne]

Try a real reset... unplug the PSU wait until all the led turn off on the board.. or 30 sec
remove the battery and use something to jump the 2 connectors of the battery holder for few seconds
the metal side of the battery do it well.

[vulcan500rider]

Thanks, MarkOne. That will be the next thing I try. ATM, I'm running a simultaneous artifact search using ATITool, and Large FFTs on each CPU core using Prime for max power consumption. My multimeter hasn't even wavered. I'm thinking the PSU is fine.

[vulcan500rider]

OK. I'm not sure I believe this myself, but I think I found the problem...

I noticed at one point that the fans on the Accelero I had fit to one of my 8800GTs weren't turning. I checked and double checked that the header was getting power, but nada. I swapped back to the original cooler, and it works fine (other than being noisy as hell, but that's no worse than the other 8800GT, I guess). Now everything seems to run without a hitch. I loaded the cards up in SLI and ran simultaneous GPU and CPU tests for max power consumption, and everything looks to be rock solid.

Problem solved...I hope. But a short in a GPU fan? That seems a little...implausible to me.

Thanks to anyone who looked, and to MarkOne for your advice. Cheers, mate! I'll update if anything changes again...for a while I was wondering if maybe my UPS was dying or something...Hmm..

[Zero82z]

It might not be a short; the fan could just be dead. You could try oiling the bearing and see if that helps.

[MarkOne]

Originally Posted by vulcan500rider

...for a while I was wondering if maybe my UPS was dying or something...Hmm..

More than often it's the battery on the UPS that die... I guess you probably know it.. what I want to say, is if you need one shop around in place they sell industrial battery, you will be surprise of what you gone save. Just ask them to give you a battery with the same spec.

[vulcan500rider]

Good thought. I think the UPS must be fine for now, though; the rig ran without a hitch all night. I do have an old UPS sitting in my closet that I may try to fix up, so thanks again.