Random lockups

[jml1911a1]

Hi everyone,

A client of mine has an HP desktop (will post specs when I get them) that, at seemingly random times, will simply lock up: The monitor will display "No signal", the hard drive light will come on as if it's writing to the HD, and nothing can be done until a shutdown is forced. This can happen whether they are working on something (surfing the web, etc: No gaming) or whether it's at idle (was left on overnight, with the case cover off, and the problem occured at some time during the night).

The heat sink on the processor was caked with dust, and I cleaned that up really well, and the problem improved, but still exists--even with the case cover off.

I checked the PS at my shop, it reads a consistant 11.7V on the Molex connectors, whether under a burn-in test or at idle. I left it on at my place overnight, and it was fine-I couldn't get the problem to show it's ugly head at all at my shop. I didn't get a chance to run a full hardware test, as I thought the issue was solved once I cleaned out the heat sink.

It feels like a PS issue (is 11.7V enough?) exacerbated by some possible voltage spikes/dips at their house, OR a heat issue, OR a HD or RAM problem.

What are everyone's thoughts?

Thanks!

 

[jml1911a1]

Edit:

It's a HP pavillion 750n, with a P4 processor.

 

[BadBigBen]

My thought: I am betting on the 'dirty power' and the PSU being a tad low... +12v (+/- 10% = +11.8v to +13.2v)... now comes a drop in the voltage below 110v AC and the above drops even further below the 11.7v you mentioned, ie. when the AirConditioner cuts in...

should be tested with a known GOOD working PSU, that produces a steady 12v line, also check the 5v rail for fluctuation and the overall power consumption versus the sustained power delivery of the PSU...

Ben

"If it works don't fix it! If it doesn't use a sledgehammer..."

 

[cbs604]

Ben, I can see you used a M$ calculator

The lower tolerence is 10.8v, but I agree with you, test it on site with a good PSU for awhile.



Cheers,
Brodie

 

[jml1911a1]

I might try a new PS, but I'm reluctant as I don't want to play "musical parts replacement". Also, if 11.7v is within spec, I'm reluctant to believe they have power problems at their house--customer is an electrician, and I'm sure he would have noticed power fluctuations.

I'll double check with him on that, and we may just have to start trying things...

Thanks, folks. Feel free to keep the ideas coming!

 

[BadBigBen]

@cbs604 - actually mistyped that, and did not proof read what I wrote, so that I actually got the whole thing wrong again... sniff... that is what you get when you are in a hurry...


Ben

"If it works don't fix it! If it doesn't use a sledgehammer..."

 

[jml1911a1]

More:

I got the problem to occur at my shop--it really seems like a heat problem, but the heat readings are only at about 112F for the HD and the CPU. I had slowed down the fans, and that seemed to force the problem, but the temps still don't seem that high to me.

Also, to elaborate further on the issue: Without warning, the monitor goes dark, the HD light comes on and stays on, none of the CD drives can be opened/closed, and holding the power button in (as to force shutdown) does nothing--I have to unplug it.

Any more thoughts on this?

 

[Jomama46]

Geez, I came here because I have the same issue.
I have an Aopen 3xs with a Celeron 1.2. Periodically it just drops out, no signal to the monitor and the mouse and keyboard dont respond. I have to reset it. It doesnt follow any pattern that I can discern.
The PS voltages are +5=4.95; +12= 12.04; -12= -12.36; and +5VSB= 5.46.

The cpu temp rises to 41C and hovers between 40 and 41. I blew it out about an hour ago and while the temp is the same I noted that the cpu fan speed has increased from 3994 to 4042. I guessing there WAS a little dust on there.
This has been a problem for about a year, I keep the insides clean and make sure the fans (cpu + 1 front panel fan) are running.
I thinking it may be a heat related problem if heat will cause these symptoms, 'cause even tho the fan is spinning faster, the heat is the same.
Any thoughts?

Jomama

 

[BadBigBen]

I would take off the HS FAN, clean off the PAD or THERMAL GREASE on both the FAN and CPU, dust off the HS FAN...

reseat the CPU, then reapply THERMAL GREASE and reattach the now CLEANED HS FAN... and monitor the progress further...

also take a closer look at the CAPS (

http://www.badcaps.com)...

Ben

"If it works don't fix it! If it doesn't use a sledgehammer..."

 

[Jomama46]

Thanks for the link Ben, that was very interesting. And at any rate I'm going to re grease my HS.

I rebuilt a few K6-233's a couple of years ago. A business had tossed them because they wouldnt POST among other things. The "other things" were dust bunnies, frozen HD (a whopping 50MB) and loose SIMMS. The most amazing to me was it was the first time I had seen HS tape. I cleaned it off and applied some HS grease and they are still running.

Jomama

 

[jml1911a1]

Problem remains!

Brand new power supply, clean heat sink.

Problem is most consistant when turning on cold--when computer hasn't been used for some time. XP will load completely, then within 2 minutes, the lockup occurs. HD light comes on, all fans go to high speed, can't open or close CD trays, can't press and hold the power button to turn off--must pull power cord.

This one is a head scratcher--I've run stress tests on it, and it does fine. Could it be a bad mobo from lightning? Bad memory? (Memory has tested fine...)

Any ideas, folks?

 

[Freestone]

Check for bad capacitors

http://www.badcaps.net/

 

[jml1911a1]

Caps all look fine--some are a bit tilted, but nothing seems like it's going bad. No leaking from the base, no bulging.

Of course, it's possible that the caps are bad even if they show no signs of being bad. Is there a way to tell if that's the case?

I'm starting to lean toward "bad motherboard", as I have no idea what else it could be.

 

[firewolfrl]

I have to say if the CPU was heat sensitive when you slowed the fan then the CPU may be your problem. and it would make sense that the CPU would fault in the first two minutes of boot...there is a hell of alot of draw on the CPU at boot and that can cause alot of SPIKE type of momentary heat...just to note.... the chipset can cause similar errors.
try this....have a house fan blow directly on the open cause at high....then walk away.
it may still lock up if the screen saver is graphic intense. on that note swap the video card to rule that out too.
I had a customer that dropped a screw on the video card ...powered up the machine ...and promptly burned a hole right through the CPU...ouch


good luck

 

[firewolfrl]

LOL spell checker works great for mispelled words.
"try this....have a house fan blow directly on the open cause at high....then walk away"
should be:
"try this....have a house fan blow directly on the open case at high....then walk away"

I should not type when I am tired....LOL

 

[jml1911a1]

Can't figure why it would be CPU overheat--I've run CPU at 100% for extended periods with no problems.

Their screen saver is slideshow of My Pictures. I ran graphics-heavy stuff, also, with no lockups.

I dunno--I might have to tell them it's likely time for a new mobo or I'll build them a new one.

 

[Jomama46]

Well it only took a month but I finally got around to cleaning and reapplying the HS compound to this Celeron and (I hope I'm not cursing myself here) there have been no re-occurrences of the lockups. When I got the HS off I could see the compound was dry looking. It has been over a week now and the lockups were occuring daily before.

Jomama