1V/Octave Voltage Controlled Oscillator

Thanks to Tom Arnold and Ian Fritz for these Suggestions For Improvement

The Scoop

Tom Arnold kept me up to date with the tweaking he did to get the oscillators where he needed them to be.

We're getting somewhere!
Playing with this VCO made me buy Electronotes finally which got me 
talking to Ian Fritz, which has gotten me to where I am now.  More in just a 
moment on that. :-)

What I tried that only helped a little or none at all.
1. Log Pair transistors (LS318s).  Didn't help.
2. Replacing U2 with the best dual opamp I could find (Ian's suggestion an OPA2134).  Helped a little.
3. Replacing the fet with a much faster one (2N4391).  Helped a little.
4. Playing with the value of the integrator cap a bit didn't move the curve much.
5. Tried to get a little more current into the expo pair (decreased R19 a tad).  Didn't help.
6. Added Franco compensation.  Helped a little but didnt have as much effect
   as expected even at large resistor values.

What helped *a lot*.
Ian got curious and started looking at the circuit.  He suggested replacing
R46 with a diode and R47 with a 10k resistor as he saw high impedence at pin
6 and a rather large switching signal. Then bumping up the integrator cap to
about 10000pf (.01uF). *bingo* With the OPA2134 and the 2N4391 in place and 
running 1v=100hz 5v=1600, we go *sharp* at 8v, or roughly 12880. This is with 
the hitrim circuit still in place and tweaking things just a little.  This is
quite tolerable as I should be able to hit C1 thru B8 no problem.  I'll have
to assume that range is really C0 thru B8 but my counter is painfully sluggish 
at anything below 50hz.  I think with a bit of tweaking of the integrator cap 
value the high end can be pulled in just a little I just dont have a wide enough 
selection of parts to try that right now.

I'm now going to start on the second VCO which I haven't modified at all.
I'll start with the R46 to diode/R47 to 10K/cap to .01uF change, then try the opamp, 
then FET, then FET and opamp and I'll let you know how those results fair. I'm hoping that 
the change will be effective without resorting to the new opamp since its 
about 6 times the cost of a TL0x2 and not available from some suppliers.


My own experiments following Tom and Ian's advice seeem to show that the
the R46 to diode/R47 to 10K/cap to .01uF change did help alot. I continue to use
the TL082 and J210 (or MPF102).

You don't want to feed the sync-in to the junction of R47 and the anode of the newly added diode so change R34's value to 20K and mount it back where it used to go. Voila when you feed in a nice fast rising wave (square or pulse) you will get the hardest sync you ever heard. Its so hard that when the frequency of the oscillator being synced goes lower than the syncing source, its amplitude begins to decay until it eventually is silent when the frequency difference is great enough. This is because it never has the chance to ramp up because it is being reset so often by the syncing source oscillator.