[SGVLUG] Calling the brightest of cal tech and other nerds!

[Bryan Backer]

Matthew,

Since you're ruled out the normal problems, how about something
from the 'grasping at straws' bucket? :)

Perhaps it's the momentary switch that PCs use for power down?
It's bad (intermittent short) or very sensative to static so
the switch is causing a power down? Depending on the
physical design of the box, the switch could be
exposed outside your case so static is affecting it.

If you disconnect the momentary switch from the motherboard
(the switch is in the case of course - attached via a header
to your MB) and instead use two jumper wires to turn on/off, do you
still experience the problem? 

No idea if it will help, but you could try your current setup but
with reset switch and power switch not connected to MB headers.
With the case open, use a wire or something to short th MB power ON/OFF
momentary switch to turn the computer on, then close the case.
Now try your various status electricity tests to see if you can
get the machine to shut down. If you can, forget my suggestion. :)
If the machine now stays up, perhaps try swapping out
power switch/headers/connectors. If nothing, try the same with
reset switch/header/connector.

For the record, I've never had a static problem like this, but
I did have a similar problem once when cramming a small form
factor MB into a wooden box and my home-made power connector
kept getting bumped - power on. power off. power on. power off.

bb

--- Matthew Gallizzi <matthew.gallizzi at gmail.com> wrote:

> On 12/22/06, John E. Kreznar <jek at ininx.com> wrote:
> >
> > > ... static electricity keeps shutting down a computer we recently
> > > built ...
> >
> > What do you mean "shutting down"?  Does it simply freeze as though
> > something is going into a tight loop, or does power get shut off?
> 
> 
> It just shuts off. I rub socks on carpet, touch front of case, system
> completely shuts off. Doesn't freeze. Last motherboard I tried would
> restart.
> 
> Have you probed anything with a 'scope?  If the power supply is
> > turning off, it must be being commanded to do so, and it would be
> > interesting to investigate the source of the command.
> 
> 
> Probed with a scope? Hm, don't think so.
> 
> I once added memory (256 KB to a Microdata 3200 -- big stuff in the
> > '70s) which turned the computer into a veritable noise detector --
> > extremely sensitive to static discharge.  Turned out to be a design
> > flaw in the CDC memory.  I fixed it by rebuilding the refresh
> circuit.
> 
> 
> I tried with a different motherboard + different RAM ... still did
> it.
> 
> 
> 
> -- 
> Matthew Gallizzi
> 


-- bbacker at mail.yahoo.com