Strange problem with cruncher that is now down?

[Johnmark]

Asus Sabertooth x58 board with a 950 CPU quit on me today. It's had been crunching and working just fine and this is the wife's machine, yesterday I did a couple reboots playing around with software and all was well during reboots, etc.

Took it down today to install a 7770 card, she was pretty dusty so I took it to the garage to blow out the dust. Got busy and finished up after about an hour so the machine was pretty cold sitting in the garage at -5c.

Plugged in the power and no joy, nothing not even as much as a beep or fan startup. The mobo power light was on though and I had pushed the mem ok button, just trying stuff. Had a good look at everything to ensure I had not dislodged anything or a piece of crap was caught in a slot or something. Nothing everything looked good. So now I'm thinking what the heck is going on, it's dead, what did I do. So I get out the manual and start reading, about 10 min. had passed when low and behold she fired up on its own accord and I got the BIOS screen saying the the memory had been reset hit any key to continue. It booted up just fine and I proceeded to uninstall video drivers, boinc in prep. for the new video card.
In the back of my mind I'm wondering if some how the cold had prevented it from starting in the first place, as everything is working fine now. Couple reboots and then I shut it down from windows. Pull the ATI 5450 card and put the 7770 in. Hit the start button and NOTHING, just as before?? WTF??

So I put the original card back in, no joy. Then I unplug everything except CPU and reset everything, look at all wiring for a possible short, reset CMOS, pull CMOS batt. then reset,tested the power switch,shorted out the power ON pins, tried a different PSU that I know works, nodda she won't do anything. The green power light on the mobo is lit, but that is it nothing else happens.

Anyone every have a problem like this? Ideas anyone?

Your input would be appreciated !!

[bluestang]

Sounds like it wasn't happy with the blowjob it got!

Other than that, what about condensation?

[OldChap]

I am going to take a punt on this: Do you ground the board whilst you clean it? to protect from static of air rushing over surfaces.

If not resolved by the time you go to bed tonight put the MB in the room/cupboard that houses your heating at home. Just to ensure that the thing is really dry....no moisture from the cleaning process.

My old rampage formula was like this somewhat. I found that when I powered it down and disconnected the psu from the wall for 24 hours 4 times out of 5 it would re-start when I connected again. I left it running until finally after a power outage it just would not try any more.

I hate turning off machines

[Johnmark]

It could be condensation but I doubt it, with the cold temps the air is very dry.

Yup, I wear a groundeding wrist strap when cleaning and working on.

I'm in the process of pulling the mobo, so if I can't get her to fire up. I'll put it on a furnace duct for the night. Wish I had a spare 1366 mobo sitting around !

[fallwind]

Condensation would occur when you brought the cold machine back into the warm house after cleaning it. Definitely let it dry overnight (unplugged) and see what happens.

[Moo]

I have a asus 775 mobo that acts like that. The only way I can get it to boot back up is by reseating the ram.

[nanoprobe]

I have a asus 775 mobo that acts like that. The only way I can get it to boot back up is by reseating the ram.

Ram was the first thing that came to my mind also.

[Johnmark]

I've reset the ram but it doesn't even post.

I'm thinking the mobo is pooched

[D_A]

Yup, I wear a groundeding wrist strap when cleaning and working on.

That will take care of the static coming from YOU but not from the air rushing across the board. For that you need to ground yourself, the board AND the air line all to a common point. Static really is a sometimes.

[BlindFreddie]

Anyone ever have a problem like this? Ideas anyone?

Yup. Twice. And I have mostly only bad news

Q9650 that's currently in GA-P35-S3 with HD7870 was originally in a nice Asus P5Q-dlx board. Cleaned it using the vacuum cleaner. To get a smaller nozzle than the open end of the hose, I'd taped a short piece of 1/2" OD clear plastic tube into the end. Not grounded. Wasn't wearing a wrist strap. Board dead as the dodo afterwards. No LEDs, no beeps, nothing.

But this was the SECOND one! Before that, the same thing happened when I used the technique on a GA-EP45-UD3R running another Q9650. On that board, I got some LEDs on power-up but not much else. Thought it was "just one of those things" until the P5Q-dlx incident.

The only good part is that only the boards seem to have been damaged - CPU, RAM, vid cards all OK.

I still use the vac. when cleaning, but at a distance from the motherboard to suck in any dust clouds that I stir up with my damp grounded bristle paintbrushes. I also use the vac hose around the fans & heatsink fins too, where it should be safe .

I also remembered that static charge buildup due to friction with moving air can be a problem around aircraft. They are always electrically grounded before & during refuelling.

BTW, the DC "ground" (0V) coming from your PSU [is probably not][Edit]may not be**[/Edit] grounded to mains earth, so don't rely on that. The metal case of the PSU should be mains earth, and this grounds the case if it's plugged in, including when it's switched off at the wall. I don't know whether using a grounded electrically-conductive fitting in the end of a vacuum cleaner hose would help, and am leaving that for someone else to try.

If anyone finds out about the viability of repairing motherboards damaged in this way, please let us know.


**[Edit]: Just checked the earthing arrangements of 5 low-end PSUs, including one in an old /AT chassis. 3, including the old /AT have the mains earth connected to both case and DC ground. On one, the mains earth is isolated from the case & DC ground, which are connected to each other. On the 5th case, mains earth is conn. to the case, but the DC ground is isolated. Bottom line? Check! (If you haven't got one, a DMM can be had for as little as $10 or so) [/Edit]

[Johnmark]

I realize that the transfer of any fluid will create an electrical differential and thus the possibility of static build up. If I was grounded to the case and air hose nozzle then no differential should occur. I just realized though that I bought a new air nozzle recently and it's handle is made of plastic(the one I had been using is all metal)! So I should have made sure my hand was on the metal part of the nozzle, which I didn't do. So I could have created a differential
I've been cleaning computers this way for about 25 years now without and problem. I usually clean the 9 PC's in the house at least once a qtr. and again no problem until now. I shouldn't say no problem, once in a while a connector gets blown off if I'm not careful.

After thinking about it more, condensation would have been the big issue, yea the humidity is low. But there is still a lot more moisture in the warm house than the cold garage. Once that board hit the warm air, then small amounts of moisture had to condense, one of those physics things

I just awoke (for night shift) so I'm going to get myself together and then take the board off the heating vent and see what happens before I go to work. Wish me luck

[Johnmark]

I am going to take a punt on this: Do you ground the board whilst you clean it? to protect from static of air rushing over surfaces.

If not resolved by the time you go to bed tonight put the MB in the room/cupboard that houses your heating at home. Just to ensure that the thing is really dry....no moisture from the cleaning process.

My old rampage formula was like this somewhat. I found that when I powered it down and disconnected the psu from the wall for 24 hours 4 times out of 5 it would re-start when I connected again. I left it running until finally after a power outage it just would not try any more.

I hate turning off machines

So I take the board off the heating vent, hook of the basics out of the case (I had tried this yesterday) and the PSU fired up, fans ran so I thought great!
Put everything back into the case, pushed start and nothing

So it looks like I have one of those boards like yours OC. Maybe tomorrow when I get home from work it will boot again in the case.
Asus warranties the Sabertooth for 5 years so I've started the process with Asus to get the board replaced.

[wplachy]

I had a problem with a new build a few years back. Everything was new, mobo, mem, CPU, GPU, case, PS. I assembled it, pressed on and nothing. I started swaping components with one that worked and no go. I then tried pulling the mobo out of the case and bingo, came right up. I RMAed the case and when the new one came installed everything and it is still running just fine after 6+ years. I never did figure out how the case could be the problem, but it sure was.

Good luck on the trouble shooting!

[mreuter80]

Not sure what power supply you have in the machine, but there is quite a difference in power need between a ATI 7770 and a ATI 5450. Have you tried another PSU?
I had once a similar issue - machine didn't want to boot up when the room was cold. After 10 minutes having the power plug in and the heater on it started alone (like yours). At that time I tested memory, graphics card and anything else - all was good. It also turned out that the PSU I had at the time was near death. After I installed another PSU couple months later (once I got it running I didn't have problems) I never had the problem again.
Just a thought.

[BlindShot]

I had a problem with a new build a few years back. Everything was new, mobo, mem, CPU, GPU, case, PS. I assembled it, pressed on and nothing. I started swaping components with one that worked and no go. I then tried pulling the mobo out of the case and bingo, came right up. I RMAed the case and when the new one came installed everything and it is still running just fine after 6+ years. I never did figure out how the case could be the problem, but it sure was.

Good luck on the trouble shooting!

How would the case cause a problem? Maybe the i/o panel shorting out.

[my87csx481]

Ok, so, I've used my air compressor to blow off some of my boards in the past. I *thought* I would be okay as long as I didn't touch the blow gun to anything. Not so sure about that now.

If you guys that actually know what you're doing could please describe precisely what you do to prevent static damage when using compressed air for cleaning, I would greatly appreciate it. (Sorry, I'm not a sparky. Never had a mentor in that area. More mechanical. The compressor is for running air tools for automotive purposes.) Thanks.

[Johnmark]

Well she hasn't started again and Asus say's send it in for RMA no hassle, no repeating myself 10 times nothing(I'm impressed) !

To those that posted, another PSU was tried, both are good 750w PSU's. I've had the same grounding issue's as wplachy mentions but it's been a few years, all I did was put electrical tape on the stand offs and I was off to the races.

As for blowing of electrical equipment. We do it all the time in an environment where a mistake cost several millions of $$. As mentioned you need to ensure there is no differential (potential) between the item you blowing off and the air gun. Think of the air rushing past the nozzle as creating an electrical charge as it exits the nozzle, that charge wants to go to a lower state (ground) and will when the energy is high enough and/or the nozzle is close enough to a component, etc..

In my case I was not grounded properly because I had my hand on the plastic handle of the nozzle. I should have used the all metal nozzle and have the other end of my wrist strap grounded to metal on the case. Make sure your compressor is oil less, as you don't want a thin layer of oil on your components.

[OldChap]

I clean my motherboards less often I guess but I have been known to wash it in the bath, Blow it off with an air compressor set up and then leave in hot cupboard for 24 hours. comes out like new

The air gun on the (filtered) compressor is metal and I ground myself on water pipe.

All this does not stop $h1t happening though, I currently have a board with a 3770K in at stock that keeps locking up. No time to find the problem and fix it though.