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

[fons at kokkinizita.net]

On Mon, Jul 19, 2010 at 02:11:19AM -0400, Tim E. Real wrote:

> You say the switch had conductance when not pressed?

Yes. Unstable resistance, changing every time the button
was pressed and released again, between 1 and 3 kOhm.

> Was the switch that simple membrane style I like to call 
>  "juice bottle safety pop-up cap"? Or did it have an actuator 
>  arm which travels when you push it?

The latter. It's a small (8mm) plastic cube consisting of
two parts that clip together, and a spring loaded actuator.

The fixed contacts are 'printed' on the inside (i.e. flush
with the inner surface), and two small pieces of copper
on the actuator bridge the tiny gaps between them. 

> Either way a switch with undue conductance is extremely rare. 

Indeed. I did not expect it ! Nor did I expect such a low Q
component in a workstation of that price. Even a simple but
real 'microswitch' would probably never fail in this way.

> I'd say it's a total fluke. Some conductive solder may have fallen inside.

There's no space that even a very tiny drop of solder could pass
through. 

> Also, are you sure it was not an unwanted stray thin conducting solder bridge
>  or wire somewhere on or near the switch, or on the board it connects to?

Pretty sure of that.

> You'd be surprised the odd things I've seen.
> Cleaning 'black dirt' from the contacts should *not* have solved it
> unless the 'dirt' itself was conductive.

Apparently it was.


Ciao,

-- 
FA

There are three of them, and Alleline.