More Nonlinear

[Sean Costello]

chordman at flash.net wrote:
> 
> Is nonlinear feedback in a filter viable?  I was thinking of
> the possibility of voltage controllable nonlinearity actually.
> 
> Perhaps something like a starved tube in the feedback path where
> the starvation is voltage controlled.  Or perhaps a filter with
> switchable nonlinearity elements in the feedback path, these
> elements may perhaps also be controlled by an external voltage.
> Or would something like this just oscillate wildly or go
> randomly chaotic too easily?

While walking around the University of Washington this past weekend
(it's interesting how thoughts are often tied to a place), I thought of
the same idea.  In some ways, I would think that there is some
nonlinearity in the feedback of many amplifiers - the Moog ladder
filter, for example, might have nonlinearities generated when the signal
feeds back to the other base of the differential pair, given enough
gain.

A clipping circuit in the feedback path would be simple, and would
create different effects based on the amount of feedback.  My idea was
to use a full-wave rectifier in the feedback path (AC coupled through a
capacitor), as this is a rather dramatic alteration of the sound.  A
paper I read on chaotic oscillations in music describes a test system
that used a full-wave rectifier in the feedback path; the result was a
period-doubling route to chaos.  I have NO idea what this would do in
the feedback of a filter, but it might be very interesting.  What would
this do to pole migration in the Moog filter, for example?  Could the
Moog filter then be used as a chaotic oscillator, given enough feedback
gain?  Or would the filter even oscillate?

The full-wave rectifier could be implemented with a single stage of an
LM3900, for full voltage control over the amount of rectification (see
Serge Tcherepnin's patents at http://www.patents.ibm.com for
clarification).  Several stages could be cascaded. 

A question:  Can an LM3900 be used as a differential amplifier, running
off of a single-ended power supply?  

Later,

Sean Costello