Works in shop: Not on site

[DTSMAN]

Have an IBM Surepos 531 point of sale computer running Win98. Failed daily due to network related issues but would not fail in house. Environment issue? Not. Loaner computers worked just fine. It is a computer with all ports on board the motherboard, so after changing everything a couple of different times (memory, harddrive, powersupply) we replaced the motherboard and it is working fine now. We tried simulating the environment by using a network cable over 300ft to push it over the max and it worked.
I know this is a long shot kind of question, but has anyone else ran into the issue of not being able to reproduce a customers problem after running their computer for weeks and taking it on site to have it go down with in 48hrs? Maybe give me a heads up for what I should be looking for in cases like ths.

Bo

Kentucky phone support-
"Mash the Kentrol key and hit scape."

 

[edfair]

Have had it multipe times with power supplies set for 230 instead of 115.
Had it once with other than IBM type where copier power supplies were clipping the sine wave.
But you have eliminated both of these.
It sounds like it could be heat related. Maybe a sensitive CPU or a trace opening.
They can be real booger-bears, can't they? FWIW, my longest ever diagnostic effort was 3 months, although it was on much earlier technology.


Ed Fair
Give the wrong symptoms, get the wrong solutions.

 

[SkipCox]

Seen it before with bad power at the site...looks like you've covered that though with other machines working just fine in the same environment.

You may have just hit the nail on the head with a marginal motherboard coupled with other marginal conditions.

Don't feel bad...this stuff can be nearly impossible to diagnose.

Skip

http://skipcox.net

 

[G0AOZ]

I'd pay particular attention to the mains electrical supply. An ageing or fatigued PSU will possible work ok with a good steady 115v (or 230v if appropriate to your country) AC supply. Put it on a fluctuating supply with electrical noise and spikes from other equipment, and it could well throw a wobbly. Future problems of this type might benefit from being run through a good quality UPS which would help to straighten out the peaks and troughs of an unstable electrical supply.

ROGER - G0AOZ.

 

[BaudKarma]

Yep, ran into it a few times. Power would be my first guess, as an earlier poster said, marginal building power plug marginal computer power supply could add up to problems.

Weirdest one was a system that ran flawlessly in the shop, refused to boot up at all at clients business. Nothing but I/O errors. This was back in the good 'ol 386 days. We went back and forth for a while, he was decently tech savvy so I took his word for it that power, air circulation, etc were fine.

I finally went out to his office to check it personally. After an hour or so of troubleshooting, I discovered that the problem was the stand that the customer had purchased for his system. When the system was standing on its end beside his desk, it refused to boot. When I laid it flat on my workbench to troubleshoot it, no problems. Probably some deal with the hard drive head going slightly out of alignment depending on the system orientation. Given the price of hard drives at the time, the client decided to just lay his system flat and live with it.

I try not to let my ignorance prevent me from offering a strong opinion.

 

[G0AOZ]

BaudKarma's post reminded me of the time some years ago when manufacturer's advice was to format and run the hard drive in the orientation it would be placed in, i.e flat on the desk or on it's side. Once formatted it was not adviseable to change its positioning. Nowadays you see drives stood on their ends, upside down, any which way!

ROGER - G0AOZ.

 

[edfair]

Yeah, that was in the days when the head positioning was independent of the platters. ST225s were particularly vulnerable. Also to platter track migration due to heat. Format them when they came to temperature. Was also when Steve Gibson had the on the fly reformatter. You could live reformat several times as the drive was heating up to get the track positioning right at the temperature it normally worked at.

 

[wahnula]

Think about the only thing(s) that was/were different in the two locations, by process of elimination it has to be one of those. If the mainboard failed onsite but not in house, what was diffferent? The temperature, humidity, network conditions and power quality are the only possibilities.