[sdiy]Touch pads?

[harry]

Allrighty!

In simple terms, each piezo sensor is like a microphone. If they are
located too
close together, a sound will be picked up by more than one at a time.

You could "gate" them together, so that only one could be active at a
time... the
"loudest" one would be assumed to be the correct one.

If they are mounted on (lets say) individual rubber pads, mounted to a
box... then
hitting one pad will couple mechanically through the box and arrive at
the other
sensors.  If one pad triggers a quiet sound (like a small bell) and the
one next to it
triggers a snare drum... guess what you will hear ;^)

Roland used a sensor on the frame that you cannot "hit" an generated a
signal that
is subtracted from the other pads... so it is a noise cancelling scheme.

Most drum interfaces have this problem BIG TIME especially if the
sensors are
attached to acoustic drums. These big surfaces can pick up a lot of
vibration
(almost as if they were designed to vibrate!).  They usually use some
kind of active
masking that lowers the sensitivity of adjacent triggers when one is
hit.

The best approach is to mechanically isolate the triggers so that
vibration does not
easily couple from one to another.  My first guess would be to punch a
hole on the
case, and glue separate rubber pads over the holes. Ther glue the piezo
to the back.
With separate pads, the vibration has to get through the rubber, through
the metal, and
through the rubber again to trigger another pad.

Interlink Electronics was the company that makes (or made?) force
sensing resaistors.
They have much less chance of interaction. My DrumKat uses this
technology... and
still has a training/interaction matrix to fix the "non-problem"

more questions? ask!

H^) harry

Alex Dickey wrote:

> > I'd use Piezo Sensors and sense impact. Almost all cheap-o
> > drum sensors use this technology. It is velocity sensitive. The
> > hard part is preventing interaction from mechanical coupling.
> > Roland cured this by summing in an extra trigger mounted in the
> > frame. Other methods work also.
>
> can you say more about this?  maybe something about the problem of
> mechanical coupling as a start?
>
> thanks,
> alex