>
> I had a problem like this once.  The problem was that the monitor was
> feeding back a voltage to the video card and causing overheating and
> eventual (days later) failure.  After replacing several video cards, we
> placed
> a fan to blow air over the chips on the card until customer could buy a new
> monitor.
>
> Don't know if this has a bearing on your problem, but it is just my
> experience.
>
>                          Hershel
>
>
>
> >A client has a PII-266 computer (ATX case), with LX mobo, S/B sound
> >card, Viper330 4Mb AGP video running win95 with lots of games. His video
> >card has already failed four times since he bought the computer about
> >year and half ago. I checked his computer, everything else is fine
> >excapt the video, when boot up, the first post is to report the video
> >memory - only 2Mb! then in win95, I can't change display setting any
> >better than 640*480 16 Color. However, there is no problem shown in
> >device manager. For test, I changed his video to a PCI card, everything
> >was fine. I thought it might be a problem with win95, but his computer
> >was taking AGP running without any problem. If it is a hardware problem,
> >it's also hard to explain why 4 video cards failed in short time.
> >
>
> ------------------------------

Sounds like you are on the right track.

Changing the Power Supply pretty much eliminates that possibility.
Aside from freakish, really rare intermittent faults, as long as the
voltages at the connections (Watch out for shorts!) to the motherboard are
within 5% of specification (including any "fuzz" you might see with an
oscilloscope), the Power Supply should be good.

Even a "killer" mother board wouldn't be too likely to affect both AGP
and PCI cards.  If you have not sent in all the dead video cards for
warranty, I suggest you try them in a test system.  Assuming the system
boots (dir a: or c: causes the drive LED to light), check them with an
oscilloscope.  If, say, the horizontal synch. signal is missing, then
check that circuit on the cable and monitor.

An intermittent open (causing repeated voltage spikes) can cause just as
much damage as an overload caused by an intermittent ground or an overload
due to a "leaky" component.  Unless you are familiar with the potentially
LETHAL high voltage and vacuum hazards, and are certain that you can safely
avoid them - then pass this information (along with the monitor) to your
favourite monitor repair person.  Good data on the problem can really
cut the cost of repairs.

I presume you have already eliminated the possibility of short circuits
caused by a loose or badly seated video card (e.g. case bent so that
the card is barely in the slot).

The only other possibility that occurs to me is a run of really bad luck.

Hope this helps!

Boyd Ramsay

[log in to unmask]

               PCBUILD mailing list is brought to you by:
                            The NOSPIN Group
                            http://nospin.com