[sdiy] Variable waveform LFO

[Don Tillman]

   > From: "Mike" <profpep at hotmail.com>
   > Date: Fri, 12 Jan 2007 18:07:19 -0000
   >
   > Looking at this, I vaguely remember a circuit for a variable waveform
   > oscillator from years ago, working on a function generator principle. It
   > depended upon a switchable current mirror, (I think), in the charging
   > circuit. This allowed the main waveforms to be varied from a positive decay
   > sawtooth via a triangle  to a negative decay sawtooth, on one output, and
   > from about a 5% to a 95% mark-space squarewave on the other. 

Sounds like mine: http://www.till.com/articles

   > Any suggestions, or is my google-fu not good enough?

Searching for "voltage controlled sawtooth" works well.  "Sawtooth
duty cycle" works even better.  :-)

There are several different approaches depending on what you care
about.  

The two-diodes-and-a pot in the VCO core approach works well, but that
isn't voltage controlled and may have a slight detuning effect on the
pitch.  So that's good for LFO's with manual control.

You could do a voltage controlled version of that, with OTA's
replacing the pot, but there would be a much greater effect on the
pitch.  So that's okay for LFO's with voltage controlled waveshape.

My approach is a post-processor on a sawtooth wave, which is a little
more work, but it's voltage controlled and is guaranteed not to affect
the pitch.  So that's good for audio VCO's with voltage controlled
waveshape.

I don't believe it's possible to implement a voltage controlled
sawtooth shaper in the VCO core which doesn't cause detuning, mostly
because our tuning standards for music are so precise.  So that's why
I went for a post processor.

  -- Don

-- 
Don Tillman
Palo Alto, California
don at till.com
http://www.till.com