MOTM

Seems simple enough conceptually. Square up the input signal with a 
comparator, feed it to a counter and timer that determine the time 
between 2 cycles, use that number to index into a memory lookup table 
representing the output voltage transfer function that you want, use 
a DAC to bring it back into the analog realm. Add an envelope 
follower cause you'll need a gate extracter, a latch to hold the 
lookup table address at the last valid value whenever the input 
signal disappears, and bingo!

The fun would be in designing some data compression tricks so that 
you would have good resolution, but could avoid a giant table.

It would still have the limitation that all P-V converters have, in 
that you have to wait a full cycle before you know what the pitch is -
 the source of notorious lag problems on bass frequencies.

Moe

--- In motm@egroups.com, "Tkacs, Ken" <ken.tkacs@j...> wrote:
> I don't know if a pitch-cv converter is the kind of thing you want 
to return
> to a 25-year-old schematic for, though, is it? It's not like a 
filter that
> might have a vintage sound; in this case you have a very specific 
result in
> mind, and accuracy is everything. I would think (could be wrong) 
that after
> a quarter century of electronic advances, we (not meaning 'me') 
should be
> able to come up with a better design, no?