[sdiy] PLL-controlled VCO

[Fredrik Carlqvist]

Hello!

I think it would be possible. You must convert the output of the VCO to
a square wave somehow. It should not be too difficult. The loop filter
will only make things interesting. You will want to control the damping
manually. Sometimes overshooting is cool, and there's nothing wrong with
glide. It could be quite interesting to experiment with these things.
You should expect the 4046 to take at least 20 VCO cycles to find the
new note with optimum damping. So for bass notes, this could take a
while, perhaps up to half a second before the pitch is perceived as
steady. Also, the loop filter will not work optimally over 8 octaves,
probably more like 4 or 5. Perhaps you will want several loop filters,
selectable with a rotary switch.

To avoid this (although I think it is probably useful that way too), you
must let the computer take control of both the control voltage and reset
timing: Let the reference oscillator (in this case the C64) also control
the current into the VCO cap. Use a one shot (555 + diode) from the
reference oscillator to reset the VCO. Adjust the control voltage to the
VCO using a DAC controlled by the C64 to adjust the VCO amplitude. So
the C64 controls the timing of the reset pulses and via the DAC also the
slope. I've had good results with this method, and I didn't even have a
proper expo converter. As long as  the VCO uses an integrator, it is not
so much trouble to do this. 

Since you probably have a expo converter in your VCO already,
determining the proper control voltage will be trivial. Hooking up a DAC
to the C64 is not so hard either. 

If you let the VCO internal reset circuitry remain intact, you can
create hard-sync waveforms just by increasing the CV above the nominal
level (not shutting down the master reset from the C64). 

Hope this helps. 


Fredrik C