To connect to a cut-down full size membrane, it would be relatively easy to fabricate a very thin piece of circuit board material and use the barrel plate and back plate to clamp it in place -- to make a pressure-fit connection to the membrane. If there was a strip along the top side and right side of the membrane, then traces on the circuit board could connect any membrane traces interrupted by the cutting, and bring necessary traces out to a ribbon cable connector.
Have you asked whether perhaps Unicomp would let you use their CAD files as a starting point for a new CAD file? You might have to agree to license the new CAD file back to them. And perhaps they would connect you with their vendors to fabricate a run of membranes -- maybe split the tooling cost in exchange for a hundred pieces.
If only...
The back and barrel plate don't cover the mylar, there is a sort of "snap-in" pressure connection. So I have done what you suggest using steel bar stock and 4-40 bolts. Unfortunately, this only works for the column circuits, not the rows. The rows are located such that an additional pressure connection is needed, but there is room for a bolt-hole only on one side. So my latest fix for
that is a pair of 1cm x 2cm flat magnets holding the membrane sections together in that location.
Anyhow, I also assumed it would be relatively simple, and in fact have the correct portion of the barrel plate cut out for my first experiment. It did not provide enough constant pressure to make reliable contact, so I moved on to the complicated method described above.
This project is currently very low on the list for me, so I have not worked on it much recently. There is also a thread elsewhere on GH describing others' work on a custom membrane.
Thanks!
- Ron | samwisekoi