MIDI for a Hammond

Robert Schrum Robert.Schrum at harpercollins.com
Fri Jul 24 18:59:00 CEST 1998

Out of curiosity, I got out my trusty 13-trace 'scope and tried it out.  ;)
(actually, I simulated it in Microsoft Excel)

The attached image time-graphs a note pressed very slowly and released very
quickly. The velocity-controlled keying voltage, noise and release bounce LFO
are shown separately, summed with the noise attenuated, and fed into 9 Schmitt
triggers with slightly different thresholds, which represent the 9 key contacts
on the 9 keyboard harmonic bus bars (labeled as the traditional pipe-organ
footages appear on the drawbars).

The main difficulty here is obviously in trying to simulate the nuances of an
extremely simple mechanical system with an electronic one.  Noise simulates dirt
and rough contact surfaces.  Attack and release slope simulates time required
for key to travel.  Differing trigger thresholds simulate differences in
tolerance and geometry of key contact wires and bars.  Release LFO simulates key
bouncing off rubber stops.  The problem is, all these things contribute greatly
to the sound of the beast.  If you could simultaneously create up to 61 of those
summed controlling voltages per manual (4th trace) with a CPU, the keying
part-count would be reduced to just 549 switching devices and maybe those
slightly-mismatched resistors if the switchers aren't inconsistent enough.

(I'll have to suggest that release bounce to Dave Amels for the next revision of
his Voce V3 module's firmware.  I already asked him for a MIDI controller
assignment to "turn the power switch off and on" (very cool and unique pitch and
gain structure changes as the tonewheels and preamp lose and regain power.)
-------------- next part --------------
A non-text attachment was scrubbed...
Type: image/gif
Size: 12052 bytes
Desc: GIF image format
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/19980724/28a933cd/attachment.gif>

More information about the Synth-diy mailing list