[sdiy] Imperfect VCO

[Grant Richter]

> Hmm.  What I'm thinking about is waveshaping using some distortion producing
> element, such as an OTA with no diode bias and a large input signal, or a 4069
> "linear" amplifier to warp the ramp.  My hope is that I can get a fairly good
> amount of distortion by subtracting the original waveform.

Remember the old "cat whisker" detectors? They work on the fact that many
oxidized conductive surfaces will have a non-linear transfer function.

You could try touching patch cord tips to highly oxidized coins. This may
sound ridiculous, but the theory is accurate.

Very much work has been put into making analog circuits linear. After many
years of research, I have found very little on making interesting
"nonlinear" circuits. Mike Firman wrote a C program that takes a harmonic
input table, and generates a Mini-Wave transfer function that converts a
unit amplitude sine wave into that harmonic series (Tchebychev method). Just
another static waveform, yawn.

Some other non-linear circuits that might be interesting:

Square root circuit using MOSFET transistor
http://www.e-insite.net/ednmag/archives/1997/112097/24di_07.htm

PAIA Sine Converter and Fuzz Boxes
http://www.muzique.com/schem/2720-14.gif
http://www.muzique.com/schem/index.html

Analog switch distortion
http://www.maxim-ic.com/appnotes.cfm/appnote_number/730