Get "hard sync" sound out of a guitar?

[WeAreAs1 at aol.com]

In a message dated 6/3/99 8:59:45 PM, costello at seanet.com wrote:

<<Speaking of, anyone have any good ideas of how to get a good "hard sync" 
sound out of a guitar? i.e.
the sound of an oscillator that is swept in frequency, hard synced to a 
master oscillator that stays
at one pitch. I suppose you could do it with a combination of a resonant 
filter, envelope following,
and distortion - I've obtained very hard-sync sounds out of a number of 
synths this way. Any
alternate ideas of how to implement this?>>

I know of a couple of excellent methods for doing this, but they are both 
pretty complicated.  First, there's the Roland GR-300 
pitch-to-period-to-voltage-to-sawtooth circuit, which tracks guitar strings 
very well (much better than any guitar-to-MIDI box you've played).  It 
doesn't have sync built in, but it wouldn't be too difficult to derive a 
reset pulse from the tracking converter.  Second, there's the very rare 
Sequential Circuits Pro/FX unit, which was a modular effects rack that had 
plug-in programmable effect modules.  One of the modules was a tracking 
oscillator with hard sync.  I had a chance to play one once, and it worked 
really well!  It had that classic Prophet V sync sound, with an envelope 
follower to sweep the synced oscillator.  You could also turn off the sync 
and use it as a tracking interval follower (tuned to a third, fifth, octave, 
or whatever).  

I have schematics for this unit, and it's quite an odd circuit indeed.  It 
uses standard parts for the most part, but it does have a couple of Curtis 
3320 filter chips.  The oscillator circuitry is discrete, fortunately.  The 
Curtis filters are configured as low and high pass filters, respectively, but 
are not part of the audio path.  They are set up to track the incoming guitar 
note and, via a clever phase-lock/feedback loop scheme, follow its pitch and 
filter all but its fundamental.  The fundamental is then squared, 
differentiated, and used to trip the reset on a sawtooth VCO.  The PLL 
circuit also outputs a rough 1v/oct, which goes to the two filters and the 
sawtooth CV node.  There's also a sample and hold that holds the last 
pitch/voltage that was played, so the note doesn't fizzle out and go off 
pitch as the guitar string dies down.  The whole mess is quite complicated, 
and probably not easy to DIY.  I'm explaining this circuit from memory, so I 
might have it partially wrong, but you get the idea, I think.

There may indeed be simpler ways to achieve the synced VCO sound on a guitar, 
especially if you don't need it to track the guitar's pitch.  Maybe a 
flip-flop octave circuit with some kind of reset, connected to a (silent) 
master oscillator?  Scott Gravenhorst has done some interesting things with 
phase-locked loops, too.  He might have some ideas.

Michael Bacich