Korg MS 20 f->Volt Converter

[Brian Towles]

> Yes, I think this is a better approach, given that the input is a perfect
> square.  It should be possible to set up a dual ramp system which does
> measurements in both parts of the wave.  To get a better square wave there
> are clever peak detectors (cascaded), I have patents from Moog and Bode
> about that somewhere.  The idea is that the largest positive peak and
> the abs largest negative peak give better information about the period
> then zero crossings do.  Just think of a fundamental with one harmonic,
> the fundamental only beeing 10% of harmonic amplitude. *YOU* would figure
> out the frequency by looking at the peaks, not at the crossings, right?

< chop, chop, ... >

this idea is built into the electro-harmonix guitar/bass microsynth.  it contains
two comparators that look for pos/neg absolute peaks respectively.  these
comparators drive a s/r flip-flop giving a nice square-wave output ( but not
necessarily 50/50 duty cycles ).  the tracking on the circuit was acceptable for
single-string guitar stuff, but sometimes liked to oscillate between octaves as
the note decayed.

-brian