[sdiy] Velocity-Sensing (was: About pots, and other things...)

[flightofharmony]

-> -----Original Message-----
-> From: synth-diy-bounces at dropmix.xs4all.nl [mailto:synth-diy-
-> bounces at dropmix.xs4all.nl] On Behalf Of ChristianH
-> Sent: Wednesday, July 02, 2008 1:56 AM
-> To: synth-diy at dropmix.xs4all.nl
-> Subject: Re: [sdiy] About pots, and other things...
-> 
-> On Tue, 1 Jul 2008 10:44:03 -0700 "flightofharmony"
-> <flight at flightofharmony.com> wrote:
-> > 3) Theoretical musing (A.K.A. Has anybody done this?): Embed a small
motor
-> > in a keyboard key for sensing velocity? I tested this out on some
scavenged
-> > drive motors and a cell phone vibration motor and it seems to work
nicely.
-> > Since the voltage induce in a winding is directly proportional to the
-> > velocity of the magnet's rotation, the generated voltage and/or current
-> > could be used almost directly. A scaling buffer would be all that is
-> > required to adapt the output to a given circuit. As a bonus, if the
motor is
-> > large enough, it would also provide tactile feedback similar to the
Moog
-> > keyboard patent which used magnets for that purpose alone. Upon release
of
-> > the key, another circuit could provide a current to force the key back
to
-> > the rest position - eliminating the need for springs.
-> 
-> If you had no springs at all - how would the key know that you have
-> released it?
-> Without a spring, the motor wouldn't generate anything on release,
-> unless you grab the key an push it up again...
-> 

good point, maybe when there is zero signal output? If the key is all the
way down no voltage is generated. Limiting the current would control the
return force  -and prevent burning out the windings while the key is held
down. A simple switch of whatever kind you desire would disengage the return
current when the key is in the up position.

Expounding further: for a more complete haptic feedback system, two windings
would be required - one to generate the CV output, and the other to provide
tactile response. This would make it easy to simulate weighted keys.

Yet further (whee!):  This would be a fun way to make a modern Player Piano,
all the keys are motorized! Especially if the dual-winding setup is used.
Yes, it is damn silly to have to provide control voltages to a keyboard in
order to generate control voltages for a synth, but it sounds like fun as
well! Maybe use it as a MIDI-to-CV converter (once a way was figured out how
to control the motors via MIDI).


-> Add to that, the motor only gives relative movement info, so you need to
-> be careful to somehow recognise "key fully down" and "fully up".
-> Otherwise any errors in integration of relative output might sum up over
-> time to either signal bleed-through on a released key, or not reaching
-> full intensity on a pressed key.
-> 

I'm not sure that need apply here (sounds like digital thinking to me). The
whole point was to have a keypress itself generate a voltage relative to the
velocity of the key, no integration or summing required, at most a buffer or
scaling amplifier on the output. Of course, this requires a "classic"-type
motor, not a stepping motor. Just a winding and a magnet essentially.
Another benefit is  that no power supply is required for the basic
operation! As for relative output, Faraday's law gives EMF = NBAω sin(ωt),
where N= number of turns, B = magnetic field, A = area of coil, ω = angular
velocity. Through the joy of Reckless Simplification, letting E =
electromotive force (voltage), v = velocity (shamelessly ignoring angular
motion for the nonce), and C = coil-related stuff (N and A), this  gives us
E = vCB which says that the voltage generated is directly related to ONLY
the velocity in this situation. Just what the original concept asked for.

The above-mentioned switch could also be used to signal which key was
pressed, along with Note On/Off messages.

-> Definitely an interesting idea, but I'm sceptical about practical
-> usability. May need quite some intensive research. But then, if it was
-> easy, any geek on the loose would have done it already :-)
-> 

Heh, practical? We don't need no stinkin' practical! Even a Cabid contrarian
and devout Iconoclast such as myself would have a hard time justifying the
practicality of ANYTHING synth- or music-related without resorting to
nebulous concepts.

Quatsch! Wir brauchen keinen Sinn!
:D

Oh yeah, a few have pointed out that Hall-effect sensors could provide a
much simpler method than this thought experiment. Yup. But this approach
also lends itself to the afore-mentioned haptic possibilities, as well as
being - in its simplest design - a *passive* methodology.

-> If you find out more, keep us informed...
-> 

k!

-> Christian
-> 

~flight
flight at flightofharmony.com
http://www.flightofharmony.com