AW: Re[2]: Pitch -> CV conversion

[Synthaholic AKA sPEW]

I am assuming that the major objection to a PLL is that the filter
adds a 'portamento' to the derived CV that might not be desirable.  I
have played with the loop filter by using a pot instead of a fixed
damping resistor.  This gives some control over the response to pitch
changes, but it is certainly not instantaneous and can cause
over-undershoot "sproing" when set too small.

I was just thinking that it might be possible to improve the tracking
by using a method similar to the GR-50 MIDI guitar box (Roland).  They
use an 'intelligent guess' by examining the dv/dt of the first quarter
cycle of the picked string.  This guess is sent as a note on message
and then quickly adjusted by a series of pitch wheel messages.  I am
not suggesting this method be used directly, but applying the same
sort of idea to a PLL system.  Suppose one designs a PLL with a loop
filter that feeds one input of a dual input summing node before going
to the PLL's VCO.  Then a guess mode circuit could watch for input
pitch changes. (like a rectified low pass filter and then sense the
change in DC output... bypass cap?).  Whenever the pitch changes, a
'helper' current is applied to the other summing node input for a
short period of time to increase the speed of response to the pitch
change.

Ok, I'll put the mushrooms away now...


On Tue, 01 Apr 97 10:32:00 PST, you wrote:

>     I think I tried that - the problem was, if there is no loop filter in 
>     the CV to the PLL VCO, the CV looks quite complex on the scope and if 
>     you just filter it to provide a DC average, then this will not cause 
>     an external synthesizer to be anywhere near the PLL VCO in pitch. The 
>     CV output must be the same as the PLL VCO CV.
>     
>     - Gene


>>      best tracking and response happens when there is *no* loop filter at 
>>      all (!) but the CV is un-useable for controlling an external synth. I 
>>      ended up with a compromise between good tracking and clean CV.
>     
>Maybe if you put the filter outside of the loop, just into the CV output 
>path,
>you'd get best tracking and still a clean CV? As the filter would not be 
>in the
>loop anymore, it could be of higher order, i.e. a better compromise 
>for audio frequency rejection vs. corner frequency.
>Just guessing, really. Have you tried it?
>     
>JH
>

- Scott Gravenhorst (Synthaholic)     www.concentric.net/~chordman

Programming: The Ultimate Computer Game. | Windows 95: The Ultimate
Unfortunately, you never win.            |      Pain in the Butt

"I didn't do it."
   -- Bart Simpson