I followed a discussion on comp.arch.embedded about prototyping BGAs, on whihc
someone mentioned a test socket which used elastomeric contacts.  I must have
translated this to 'rubber' in my mind.  My first comment was aimed at the
rubber used to connect the LCD and pcb, not the keypads.  I have only run across
it a few times, and manufacturers use ribbon cables mostly.  I did fix a watch
recently wich used this rubber strip to connect the lcd to the pcb.  At any
rate, when this message popped up, the closest thing to his description that I
had dealt with was the lcd-pcb rubber connector.

The message about the elastomeric test socket can be found here:
http://x24.deja.com/getdoc.xp?AN=384414864&search=thread&CONTEXT=936109849.97609
3336&HIT_CONTEXT=936109055.974651538&HIT_NUM=126&hitnum=6

(yes, my mail program autowraps, so you'll need to copy both parts (without
spaces) into your browser...)

-Adam

Tom Handley wrote:
>
>    Adam, I really don't have experience here. The conductive rubber that
> is typically found in keypads/keyboards, etc, has too much resistance. As
> far as BGAs, I could look into it but I was hoping that someone would
> speak up and give the details. My `gut feeling' is that there is a good
> reason this technology is not being used in test and programming sockets.
>
>    - Tom
>
> At 11:07 AM 8/26/99 -0400, Adam Davis wrote:
> >I've heard they have test and programming sockets at a huge expense.  I've
> also
> >heard of a less expensive methed, but requires some specialized materials:
> >
> >Think of a rubber pad, about the same width and length of the BGA, and about
> >1-5mm thick.  The rubber pad is conductive vertically (through the pad)
> but not
> >horizontally(along the lendth and width of it).  Place the pad on a pcb
> with the
> >bga pattern on it, and press the bga against the pad.  Assuming it's all
> lined
> >up, you're all set.
> >
> >I've seen strips of rubber which are used in older calcs to connect the
> pcb with
> >the lcd.  I assume the rubber pad has the same pronciples...
> >
> >-Adam
> >
> >Tom Handley wrote:
> >>
> >>    Wagner, I was thinking of those DIP-SOIC adapters where you have to
> >> solder the SOIC. I think they were around 6-$10 depending on size. This
> >> is single qty.
> >>
> >>    If you are talking about sockets, that's another story. They are
> >> expensive. Your idea about replacing a spring contact sounds interesting
> >> but it depends on the application. It seems to me that there would still
> >> be too much resistance for many applications. The expensive sockets are
> >> intended as universal test sockets. The conductive rubber approach would
> >> limit it's application. This is really outside of my expertise but I was
> >> wondering what they use in Ball Grid Arrays? Can those `conductive bumps'
> >> be applied to a socket and at what cost?
> >>
> >>    - Tom
>
> ------------------------------------------------------------------------
> Tom Handley
> New Age Communications
> Since '75 before "New Age" and no one around here is waiting for UFOs ;-)