resonance with a svf

patchell patchell at
Fri Aug 11 00:06:57 CEST 2000

    One thing I have always wanted to try is based on a circuit based on
the quadrature sine wave oscilator that you will find in the Analog
Devices NonLinear Circuits handbook (page 80 in my copy).  This
oscilator is basically a state variable filter.  If you look at the
amplifier that controlls the resonance, it keeps the thing oscilating by
balancing positive and negative feed back.  When the voltage exceeds a
certain amplitude, the zener diodes begin to conduct forcing negative
feedback.  When the amplitude is low, the zeners don't conduct, and you
only have positive feedback.  In the Nov./Dec. 1980 issue of Polyphony,
you will find the same circuit, except implemented with CA3280's instead
of analog multipliers.  The article was written by some strange
character. :^)

    If you use an OTA to control the resonance, and the OTA is
"inverting", if you put a large resistor around the OTA, this resistor
will provide positive feedback when the OTA is at very low Gm values.
This might be enough to make it start to oscillate.  By using some other
non-linear elements, you might be able to control the oscilations so
that it doesn't run away.

    Anyway, just some rambling thoughts.

modlar at wrote:

> Hi all !
> Thanks for all the good ideas to solve my problem.
> I am just working on the filter with the 3080 (the LM does
> not work, so I gave up).
> My next question (Sorry, only questions...) is :
> is it possible (theoretically) to design a state var filter,
> (like the formant filter, for example) that is able to
> "self-oscillate" ? Some months ago, I remember that someone
> said, that a SVF is not able to produce self oscillation.
> Is that true ? Or, if not, what is to change ?
> My filter design looks like the formant filter, but with
> 2 074-OPAmps instead of the BF245's.
> How can I produce self oscillation (squeeeeeeeeeez!!!!!) ?
> Changing of R19 (decrease to 0 Ohm), R18 (decrease to 0 Ohm)
> does not work, and amplification of the resonant pathway (before
> P5, using an OPAmp) does not help either.
> Arhhgrggrgrg !!!!!!!
> Any help ?????????
> --
> Sent through GMX FreeMail -

* Visit:
*If you didn't buy a home in Santa Barbara,
* You didn't pay enough!

More information about the Synth-diy mailing list