[sdiy] Minimoog clone
David G. Dixon
dixon at interchange.ubc.ca
Sun Jul 11 18:25:49 CEST 2010
> A keyboard is called "matrixed" because it electrically appears to be a
> grid of switches. You know when a particular key is pressed when a
> circuit is completed when a circuit goes from the column to row wire. If
> too many keys are down at once, false keypresses or "ghosting" can be
> created. People using gutted computer keyboards to create arcade
> controller interfaces had lots of problems with this and that's why
> special non-matrix gizmos exist for that application. Careful design of
> the matrix can lessen, but not prevent the occurence of ghosting. From
> some mental doodling, a matrix can be made into a non-matrix with an AND
> gate for each key, but it's still prone to ghosting.
Hey David,
I built (a slightly modified version of) Ray Wilson's matrix scanning
circuit for some matrix keyboards I ripped out of an old organ:
http://www.musicfromouterspace.com/analogsynth/KEYENCODERPCB/KBDMATRIX022005
.html
This circuit is low-note-priority and generates 1V/oct CV, gates and
triggers (although the 12V gate and trigger circuits shown in the third
schematic are unworkable -- just keep them at 5V and you won't go wrong).
It works very well, and I've never experienced "ghosting". In fact, until
you mentioned it, I'd never heard of it. The circuit latches onto the last
low note played, and doesn't scan higher until that note is released, so I
don't see how things could go terribly wrong.
Of course, perhaps you want a buss keyboard just for the authenticity of it,
but at 1000 scans per second, you can play as fast as humanly possible and
you will never hear any glitches.
More information about the Synth-diy
mailing list