Troubleshooting Random hangs, etc.

[Endaar]

Hi All,

I'm thoroughly stumped with the below and would appreciate any thoughts or advice as to how to proceed. The system in question:

Gigabyte EP45-UD3P motherboard
4GB Corsair Memory
(2) 10K 36GB WD SATA HDDs (RAID 1 using Intel controller on m/b)
(1) 500GB WD SATA HDD (using the other m/b controller)
(2) Plextor SATA DVD-R/W drives
Antec HE550 PSU (3 12V rails @ 18A each)
SB Audigy Gamer
See below for video card
Vista Business x32

About a week ago, I had a couple of power blips/surges/whatever at home. Despite the PC being on a surge-protected UPS, it apparently took some damage as the next day when I powered-on the system, the monitor would not come out of sleep mode. I tested the monitor with another system and verified it's working fine.

Eventually, after a bunch of reboots, I managed to get some video. There was a lot of garbage on the screen, and when I checked device manager, the video card (an EVGA 8800GTX) showed an exclamation point. I tried to "uninstall" the card from Windows, and having done that, could not get any video out of the system again, including during POST. I was pretty certain at that point the video card was bad.

EVGA would not honor the "lifetime" warranty on the card, so I purchased a Gigabyte GTX 260 OC. After swapping cards, I had a hard time getting the new driver to install. It took a few reboots into both safe mode and normal startup, but windows eventually sorted itself out.

I ran the nVidia stability test from within the video card's control panel, and the system eventually hung. Since then I've been dealing with random hangs and a frequent refusal to finish booting into Windows. I don't see any pattern to those hangs.

Since the only thing that changed was the video card, I focused on the driver. I had downloaded the latest 185 driver, but in searching the web found that some people had issues with it and reccomended 182. I removed the 185 driver, ran Driver Cleaner, and with the system just using generic VGA drivers, things were OK. Immediately after installing the 182 driver I began to get rundll32 crashes and the system went back to being unstable.

So now I'm thinking there was a conflict between the various drivers and the only way to fix it is a clean install of the OS. I started to do just that and had the install hang when it should do the first reboot. I manually rebooted the PC and the install eventually finished. Again, things seemed OK with standard VGA drivers.

I installed the nVidia drivers and the monitor went into sleep mode after the reboot. This happened a few times before I went into safe mode and had video again.

I figured I'd give XP a try just to rule out anything strange with Vista, but the XP install bluescreened just before the F8 license agreement page.

After another reboot, I was back in Vista but the video card drivers failed to load because of a missing file. A CHKDSK turned up a bunch of errors on the HDD. Perhaps related, the RAID controller was frequently "verifying" the RAID 1 volume. After running CHKDSK, I was going to reinstall the video driver but the PC hung before I could do so.

My last thought was power. The card requires a 500W PSU with 36A on the 12V line. The Antec is 550W with (3) 18A lines, so it seems like it should be OK. I disconnected everything non-essential, and still have the same problem. Since the system doesn't randomly reboot and the hangs happen at anytime, even at idle, power seems an unlikely problem, but who knows...

Anyway, that's where I'm at. I don't have spare parts to start swapping out, and am at the point now where I'm concerned the motherboard took some damage along with the video card.

Frankly, I'm out of ideas...any help would be greatly appreciated.

Thanks!
Endaar

[zapionics<alt>]

It seems pretty unlikely to me that these hangs and blue screens co-incidentally started happening at the same time as the power blips. I think its more likely they're related and the UPS has failed to regulate the power surges sufficiently. A good UPS should do a couple of things; provide backup power to allow graceful shutdown during power loss and also condition the incoming power as well as regulate voltage spikes. It may have failed under some extreme condition.

The very first thing I would do if you havent already is clear the CMOS. This is an easy thing to do, only takes a couple of minutes but can fix many problems! Just follow the boards user manual regards the procedure (probably short some pins or remove battery) and then reload your defaults in the BIOS. Take note of any special settings you have in the BIOS prior to clearing it.

Next, its also worth testing the memory sticks, since that is pretty easy process as well. You can download MEMTEST which is a great testing program for DDR, just checkout their website.
If the board supports running with only 1 stick then you could also try one stick at a time to try an isolate any faulty sticks.

Failing that, I think its possible the board *may* be damaged or the power supply, but I'd recommend that you check out the boot drives, since thats the next likely/easiest thing to check and the behaviour you describe could be related to corrupted disks.

Can you put the drives in another working system and run some diagnostics on them? Or otherwise take them to a shop for testing. I think the manufacturer, WD, may have some tools you could download to do diagnostics on the drives, have a look at their website.

[edit] and I was just thinking, if you have another spare working disk then you could put that in to test the system. If all is well that would point to a problem with the disk array, either one or both of the drives.
If you happen to have a spare power supply lying around (some do) you could try that as well to determine if the power supply is the problem.

[Endaar]

Thanks for the suggestions. Unfortunately I haven't made any progress.

I've cleared the CMOS with no change. I cannot boot into MEMTEST - it hangs before getting to the actual test screen. I've tried with each of my two DIMMs individually with the same result.

I tested the PSU with a voltmeter and all the voltages are where they should be and are extremely stable. I do have a spare PSU but it's not big enough to power the video card.

I didn't get around to checking the disks, but since I can't boot into MEMTEST, I think it's safe to say there's something wrong with the motherboard, memory or CPU.

Any other ideas?

Thanks,
Endaar

[Endaar]

Well, I think I've established that the motherboard is bad.

I was able to borrow some DDR2 DIMMs from work, as well as another E8400 processor. Using either or both of those good parts still does not resolve the problem. I can't see how it could be anything other than the motherboard.

Anyone have any last thoughts before I decide how I want to deal with fixing this?

Thanks,
Endaar

[zapionics<alt>]

Originally Posted by Endaar

Well, I think I've established that the motherboard is bad.

I was able to borrow some DDR2 DIMMs from work, as well as another E8400 processor. Using either or both of those good parts still does not resolve the problem. I can't see how it could be anything other than the motherboard.

Anyone have any last thoughts before I decide how I want to deal with fixing this?

Thanks,
Endaar

Hmm, it does seem likely its the board given you have replaced video card, memory and CPU. The disks still concern me but I would have expected it to boot to MEMTEST even if they are corrupted (how are u booting MEMTEST?).

Its not at all unusual for a boad to be damaged by power spikes, so this does seem reasonable assumption now. Sometimes you can see damaged components on a board (such as burnt or melted IC's) but not always.

Testing the PS with a voltmeter is not the same as testing under load, I still think its worthwhile trying the other PS before u go spending big, you can remove the video card - doesnt the board have onboard video option?

Also, u are able to boot into BIOS?

[Endaar]

I'm booting MEMTEST off of a CD created with the latest ISO download.

I unfortunately don't have access to another PSU that's powerful enough to drive the video card. Best I can do is pull the card and test with a low-end one that doesn't require separate power. The MB does not have onboard video.

Yes, I can get into the BIOS.

Thanks,
Endaar

[zapionics<alt>]

Originally Posted by Endaar

I'm booting MEMTEST off of a CD created with the latest ISO download.

I unfortunately don't have access to another PSU that's powerful enough to drive the video card. Best I can do is pull the card and test with a low-end one that doesn't require separate power. The MB does not have onboard video.

Yes, I can get into the BIOS.

Thanks,
Endaar

You said earlier that system hangs when trying to load MEMTEST. Does sound more and more like the board got badly hit. Still, power supply is between the source and it so its worth excluding as the culprit. If the board continues to misbehave with the other power supply then that is the final proof you need.