[sdiy] How does the JX3P VCA work?

[rburnett at richieburnett.co.uk]

Hi Gordon,

The VCA in the JX-3P only uses a single pass transistor between the 
attenuated source and the op-amp.  I've never seen the other arrangement 
with two matched transistors before.

I can't give you a detailed explanation of how the JX-3P VCA works other 
than to just say that it uses the transistor in the small-signal region 
where it behaves approximately like a voltage controlled resistance.

What I can tell you is that it distorts pretty badly because the 
transistor does not respond equally to the audio signal in the positive 
and negative directions.  It becomes progressively more non-linear for 
larger audio signals.  So you get significant amounts of odd and even 
harmonics at the output.  It is also very noisey due to the effect of 
the low-ohm resistor forcing up the noise-gain of the operation 
amplifier.  And finally it exhibits DC control feedthrough, because the 
base current from the CV input eventually goes into the op-amps 
virtual-earth.  I'd say that the VCAs in the JX-3P are probably just 
about as bad as they could be!

The one used in the feedback path of the VCF to control resonance is 
less critical because as long as something vaguely resembling the output 
is fed back to the input it will increase the filter's Q factor and 
achieve it's aim of controlling resonance.  Most of the horrible 
distortion products get attenuated when they pass through the 4-pole 
low-pass filter.  But this cheap VCA is definitely responsible for the 
JX-3P's strange "resonance burst around the zero crossing" behaviour in 
the filter output waveforms.

The VCAs used for each voice's amplitude control are quite nasty though. 
  I tried replacing both VCA's in one voice of my JX-3P with a dual VCA 
made from LM13700 and it improved the sound immensely.  It made the 
JX-3P sound much brighter and lively, very similar to the sound of a 
Juno 106.  It also seemed to reduce noise too, but analogue JX-3P fans 
will probably say it's meant to sound like that! ;-)

I made a PCB for mod'ing all 6 voices but never got round to completing 
it.

-Richie,


On 2014-01-15 10:32, Gordon JC Pearce wrote:
> I was looking at simple VCA circuits (I want popless muting for some
> communications gear, so a slightly ramped VCA would do but distortion
> isn't a big problem), and ran across this:
> 
> http://urekarm.tripod.com/synth/jx_vca.pdf
> 
> How exactly does it work?  I would have thought that increasing the CV
> would turn on both the pass transistor and the shunt transistor.
> There's obviously something clever that I'm missing.