[LAD] [Somewhat OT] Strange failure mode of a PC

[fons at kokkinizita.net]

Hello all,

Early this week one of the three 'rendering' PCs of the WFS
system in the Sala Bianca failed. It just appeared completely
dead and didn't even try to boot when the power button was
pressed, but the standby power (for the network interface)
was available.

I suspected the power supply, so ordered a new one which
arrived two days ago. Installed it and things worked again.

But now comes the interesting part. While installing the new
PS, I also disconnected the wires to the power button, and 
to test I just used a screwdriver to short the two pins that
normally connect to it. But when I reconnected the power button
the PC switched off after a few seconds. So it seemed as if the
power button was permanently being pressed. I again installed
the old PS, and things worked as long as the power button was
not connected. 

Measuring the power button switch with a multimeter showed
an unstable resistance value of between 1 and 3k while it
was not pressed. So I removed the thing, which turned out
to be a cheap miniature switch, a little cube of around 8
mm size. I opened and disassembled it, and noticed that the
contacts had some black dirt on them. Cleaned with aceton
and reassembled, and things worked perfectly again. 

What I don't understand is how the contacts got so dirty. 
If a resistance of a few kOhm is enough to make it look
as a closed contact then it can't be handling large currents,
so there should not be any arcing. And the construction of
the thing is such that it is virtually closed, no dust or
whatever could ever creep in. 

Still it's quite sobering that this cheap 0.30 Euro thing
was capable of bringing down a 1600 Euro workstation...  
Who would suspect a switch to fail in this way ?

Ciao,

-- 
Je veux que la mort me trouve plantant mes choux, mais
nonchalant d’elle, et encore plus de mon jardin imparfait.
(Michel de Montaigne)