PC Keyboard rewiring

I am having the following problem shown in this diagram, and have a
solution that I think might work - but I am not sure.
Here is the diagram:

http://img207.imageshack.us/img207/1161/kbproblem4bb.png

My problem is that I am trying to build a music controller -
essentially a bunch of sample triggers - and this is most cheaply done
via keymappings, which the software (Ableton Live 5) supports. But, to
get the keys I need, I will use almost every key and its shifted
complement.

This has led to two issues:

One, I can make EITHER the shifted keys work OR the non-shifted keys,
simply by activating the keyboard matrix's shift key and then wiring
spst switches to the other mappings. Alternately, I can of course leave
the shift key off, and wire stuff that way.

Two: The other issue is if I attach the shift leads to two switches,
each switch will turn on ALL keys hooked to the switch - enabled wires
- I guess the current flows through the whole circuit, so I was
thinking diodes might be the answer.

I even though about using TWO usb keyboards, and wiring one for the
shifted and one for the non-shifted keys. Either way, though, I cannot
permanently wire the shift key in the ON position, and so I still must
solve problem #2 in any case.

I'm very new to electronics, though so this might be off base. I know
there HAS to be a solution - because obviously, the keyboard has the
ability to do this prior to dissassembly!

If anyone can suggest another solution - like a way to send the keys
from a custom-built controller that sends the ASCII codes, or the scan
codes - I'd be interested in other options as well. One other
potentioal snag with my plan is that if I press a button to a shifted
character and another to a non-shifted char, BOTH chars will be
shifted, unless I can somehow figure a way to send the signals
alternately.

If you want to see the controller concept I am building, you can see it
here:
http://www.midibox.org/forum/index.php?topic=6402.0

.

Quantcast