odd, intermitent issue

[imperialreign]

Throw me some ideas here, been dealing now and then with an odd, dodgy issue:

on occasion, during POST the motherboard will respond something along the lines of "bad or incompatible checksum" and then proceed to check the HDD, floppy and USBs for a BIOS version to revert to - typically a hit of the reset switch, and the borad will POST normally.

But then, on occasion, I run into an issue booting into WIN - either it resets during the GUI screen, or it'll reset when it gets to "loading personal setting" - on very rare occasion, the system will reset long, long after you're entered the boot session; and other times, not at all even with the rig running 12+ hours. Othertimes, the SYS will lock up, but this only tends to happen when gaming.

I thought, perhaps it's a bad OC, and reverted back to 3.4GHz with known good settings - still happens intermitently. I've loaded BIOS defaults, and run everything stock, and it'll still act up now and then . . .

I considered a corrupted XP install, but that wouldn't explain the hiccups during POST; I've run memtest as that seemed to be the next usual culprit - no issues there . . .

So . . . possible bad / failing PSU? Although, all voltage sensors reported through software on the mobo seem to check out alright, I'm not eliminating the possiblity of the unit as failing, especially considering the rather random inconsistent nature of these crashes and failures . . .

Possible corrupted BIOS or a corrupted BIOS chip itself?

What other ideas can y'all throw at me on this one?

 

[spearman914]

Theres a possibility that it is your psu. But then it wouldn't make sense if the sys lock up. Can you replace the bios chip?

 

[erocker]

I would try a new flash on the motherboard to see if that corrects the problem since it's the easiest to do, plus it sounds a lot like it could be partially corrupted. I think you are also correct by assuming that it may be the PSU. When PSU's degrade, generally it starts becoming noticable at first during power on, as ripples are more likely to occur. There's no real way of testing this short of jamming probes into your wires. Good luck.

 

[imperialreign]

Theres a possibility that it is your psu. But then it wouldn't make sense if the sys lock up. Can you replace the bios chip?

not on this one - the BIOS is a hard solder (rare for ASUS), but it does have a back up feature, IIRC, that will either load the shipped BIOS, or search for a compatible ROM to flash - don't mean squat if the BIOS chip itself is toast.

I would try a new flash on the motherboard to see if that corrects the problem since it's the easiest to do, plus it sounds a lot like it could be partially corrupted. I think you are also correct by assuming that it may be the PSU. When PSU's degrade, generally it starts becoming noticable at first during power on, as ripples are more likely to occur. There's no real way of testing this short of jamming probes into your wires. Good luck

I was thinking I might as well do that later as well - flash the BIOS. The P5E3 is still rather new, I'm sure there's a more recent BIOS available. Funny thing - this happened before about a month ago, I figured it was a sketchy BIOS, so I flashed to the most current, and haven't had an issue until this last weekend.

If it still does it, I guess I'll have to break out the DVOM and run down the PSU - I haven't had an issue with this Hiper . . . yet; but I know quite a few other's have. S'all good if it's bad, anyhow; I intended to replace it one I snag a 70x2 in the next few months (something beefcake enough to handle two of them in Crossfire and still have head room - prob a 1KW or better unit).

 

[Dia01]

I've had this problem with ASUS boards in the past (being the notoriously bad 780i). I always found that is was my RAM especially when running with SLI Memory enables but I could never really narrow it down. I was using OCZ DDR2 1066 4ea X 1GB sticks, but by the looks your using DDR3. Don't know, try a bios flash like erocker said.

 

[Dr. Spankenstein]

Like Dia01, I was going to suggest you check the stability of your RAM with a few passes of Memtest.

If you get the "all clear", maybe a bump to the NB volts.

 

[imperialreign]

Like Dia01, I was going to suggest you check the stability of your RAM with a few passes of Memtest.

If you get the "all clear", maybe a bump to the NB volts.

they've passed - every time this issue crops up . . .

I reflashed the BIOS with v1109 (nothing newer) - which it was running before, thinking maybe the last BIOS install was partially corrupted, or I skewed something OCing the board

not sure on NSBv - even at my reduced OC right now (3.4Ghz), NSB is still set at 1.41, which is a bit above where it needs to be


waiting game, now after this BIOS flash . . . if it starts acting up again, I'll have to get out the DVOM and start checking the PSU.


Damn, I hate these phantom issues, though - the kind that mimic other components going bad.

 

[imperialreign]

2 days now, no issues since the re-flash :shadedshu


not sure whether to be happy about that, or leary . . . if I've got hardware starting to fail, I want to be able to know now!!

 

[lemonadesoda]

Small price to pay for OC'ing...

Also...

1./ Stip the mainboard of all attachments
2./ Dust, wipe down
3./ Look carefully at EVERY mosfet and EVERY capacitor
4./ If "brown lacquer" deposits on or near solder joints, this is a sign of overheating, tired mosfets or failing caps. All you can do it live with the fact that you pushed the MB a bit too hard... and need to UNDERCLOCK to keep the thing stable now, OR
5./ If you enjoy working with a soldering iron, replace the mosfets, and replace the capacitors. But his is a timeconsuming and risky exercise. Do it for fun only. If you NEED a working PC, get a new mainboard.
6./ And "value brand" PSU can be considered EOL after 3 years.

 

[imperialreign]

Small price to pay for OC'ing...

Also...

1./ Stip the mainboard of all attachments
2./ Dust, wipe down
3./ Look carefully at EVERY mosfet and EVERY capacitor
4./ If "brown lacquer" deposits on or near solder joints, this is a sign of overheating, tired mosfets or failing caps. All you can do it live with the fact that you pushed the MB a bit too hard... and need to UNDERCLOCK to keep the thing stable now, OR
5./ If you enjoy working with a soldering iron, replace the mosfets, and replace the capacitors. But his is a timeconsuming and risky exercise. Do it for fun only. If you NEED a working PC, get a new mainboard.
6./ And "value brand" PSU can be considered EOL after 3 years.

yeah, I understand that on the PSU - I'm kind leaning on that being the primary culprit as it's only a year old, and most user reviews I've read people start complaining of issues after a year - if it makes it that far. No biggie, though, it's help up strong under my iron fist, and I intended to replace it one I got hold of a new GPU.

As to the rest, though, I find that somewhat doubtful (although I should probably look). I've never had an ASUS board, especially a high-end ASUS board, start suffering from PCB component failure within 2.5 months of it's build date - and most of their high-end motherboards can withstand some insane abuse . . . not saying it isn't possible, only that I find it doubtful. At the very least, though, I've been planning on pulling the chipset cooler off the board this weekend and making sure I've got good contact between the cooler and the NSB, SSB and VRMs - especially over the VRMs and mosfets near the CPU socket with this board.

 

[panchoman]

i get it a lot, it doesn't affect anything really.. it usually happens after a bad oc, but every say once in a month or so i'll get it on a normal boot. it doesn't do anything to my system though.

 

[imperialreign]

i get it a lot, it doesn't affect anything really.. it usually happens after a bad oc, but every say once in a month or so i'll get it on a normal boot. it doesn't do anything to my system though.

see, that's kinda what I find odd is that this issue is opposite what one would expect . . . I can set the BUS at 400, lock the DRAM at 1600, run the CPU at 3.6GHz @ 1.45v - and it'll go weeks without an issue, through numerous boot cycles, and then all of a sudden it starts acting like how I described above - backing the clocks down doesn't typically resolve the issue, and occasionally, running default settings still equals problems . . .


but, if I go and re-flash the BIOS, everything goes back to normal again for a week or two.


I swear, my rig is trying to drive me insane