Reboots/No BSODs, Completely random/intermitent

Judas

This is totally baffling me.

Heres what happened:

Did a clean wipe and a fresh install of this machine:

specs:
Gigabyte GA-7VT600-L
AMD Athlon XP 1900+ (socket A)
Radeon 8500LE AGP (Swapped for x800xl/nvidia Geforce 2 MX400 and other cards)
Tried 256mb/512/2x256/ combination ram @ various speeds (OCZ 512 PC3200 Wouldn't post)
15Gb Maxtor ATA133 2mb cache HD
LG 52x24x52x CD-RW Master (tested unplugged)
Generic 50x CD-Rom Slave (tested unplugged)
Enermax 300 watt PSU (tested with alternative)

Any range of monitors/keyboards/etc.

Even did the strip down test, (ram/videocard/cpu/motherboard/HD).

But heres the little backstory on it.


After doing a fresh install and doing all the updates, running it for hours, rebooting a few dozen times at least. There was no problems, no crashes, ran through burn in test with no errors.

The dilivered the machine, hooked it up, hit the power button and enless booting loop. No BSOD, no nothing.

Uplugged the keyboard, and it posted find about 4 or 5 times. Plugged in a USB keyboard, and no boot, unplugged it, no post. And then things just got worse, the machine would boot up as long as nothing was plugged in, everything worked flawlessly in windows, but loading windows it would reboot.

Well much work later, i did some more diagnostics test, running through a few things just to see if the damn thing would post. But it's completely random and intermitent.

Sometimes just as i figured i'd found the problem, it would do it again.

Did a fresh install again without updates, and would do it again anyways.

Well took the damn machine back to the shop and have been screwing with it ever since.

I did however get it to finally display a BSOD (after getting into windows and then while shutting down)

PFN_LIST_CORRUPT

Looking through the possible causes, the most likely thing would be memory. Yet i tried, AGAIN, 2 different memtesters, for hours, not a single error, even tried loosening the settings to the max. Still no change in booting, no error on any burn in tests and diagnostics tests. Nothing seems to be making sense.

One thing though i must add, the client is a heavy HEAVY smoker. I mean in 6 months, the fan blades on any fan in the system is litterly so coated with tar and smoke and general crap that the fans will litterly come to a hault.

Anyone at all run into this issue before? anyone have a clue?

I've also, just to make sure, yanked everything appart, tried to clean it up as best i can with 100% rubbing alchol. Put fresh thermal paste on, made sure the heatsink was making full contact. Check the temperatures, everything is temperature wise, below typical machine temperatures. And once again, While in windows, absalutely no program hitches and errors. (well aside from the installation of HP ALL in one printer which seems to always have thier own issues)

Dyre Straits

Judas, have you tried a different PSU in it yet?

EDIT: NM..I see you say 'alternative'. Is that a better alternative?

It almost sounds like something's loose .... any bad capacitors? Bulging caps?

PangingJr

try different motherboard if you have one in the house, and see if all things seems to be better than before?
otherwise, reflash the motherboard's BIOS with a newer BIOS file version and one that is not beta. and then try again.

Judas

The PSU i tried with (the alternative) is the OCZ GamerXStream 600 watt (which i had been using in a much much more powerful system previously with no issues) the results i identical.

I've also tried flashing the bios, even incrementally to see if that would resolve the issue. I've finally made it to the last revision bios and every flash has gone successfully. But the problem still occurs.

I've since replaced the Hardrive with a known good on... Replaced ALL the cables in the machine, replaced every conceiveable part except the motherboard. And the problem seems to still excist.

However, I just found some 2nd hand PC2100 DDR ram that i've popped into the machine, but i can't be sure if the problems resolved (i haven't had a reboot yet, but hey, i've got 12-20 reboots and then a crash)

Currently got my fingers crossed. I'm guessing a memory/motherboard/via chipset incapability? (considering that the OCZ Value 512mb PC3200 ram wouldn't even post even though the motherboard support PC3200 right out of the box)

And unfortuneatly, i can't even get in a socket A motherboard.

I've also got a coolermaster 500 watt PSU on the way, i can't even get in any DDR that's sub PC3200 speeds.

Dyre Straits

Sounds like an overdue motherboard upgrade is needed.

Judas

Well

after tearing it all appart a 2nd time, and then putting it back together, it only did the reboot crash once initially... since then i must have rebooted it 200 times in the past 5 hours... Each time no issues, no oddities.

And just a little while ago, hauled it back over to the clients house, plugged it all in, and fired it up without a hitch. I don't get exactly what it is, but i had warned them that if the symptoms return, that because it's intermitent, they may return, that there isn't much of any other options aside from replacing the motherboard and other hardware that would have to be changed with it. But so far so good.. no issues... solid. WEIRD

Dyre Straits

If you do have to take it back, you might try running it with the mobo outside the case....just to make sure it wasn't intermittently shorting against something.

Good luck!

Judas

i tried that as well

but it did the same thing at the time... i'm convinced that either A: the motherboard itself is failing, or B: due to the smoke and the chemicals based in it, and knowing how heavily they smoke around the machine, that the residue collected on the board is actually creating a circut/short.

MartynGB

Had a similar problem a few years ago at a coal merchant. Suspected that coal dust was giving intermittent short circuits on th MoBo. Rebuilt the ystem in an industrial case with filtered air supply & door over the removeable media drives, no problem after that. Remember to give user instructions on how to clean and replace the filter!

H3X4D3C1M4L

I had similar issues.... what I did was change a jumper on my motherboard so the mobo got the S2K bus timings strapped from the proc itself rather than the mobo and it made things alot more stable but I'm not sure that your mobo has that problem (and it only plagued me when I ran a 3000+, never on the 1900+ I had for some reason)

Judas

IO had that issue with a Abit board as well, hardware strapping had to be done to reolve the higher 2000+ xp models.

But these users don't want complex case with filters and such.... i had suggested it.... but hey it's there choice...

H3X4D3C1M4L

Sounded alot like hardware strapping to me

I also noticed that only happened with SATA drives and never IDE drives... it was weird.

I could have SATA drives installed but couldn't use them or I'd BSOD fast.