[sdiy] Change response from Lin to Log to Lin to Exp?

[Justin Owen]

>>> A diff-amp across the VCA?  That would make the blue curve look like the green curve.

>> I assume you mean a single input Differentiator - not a dual input Differential?

> No, a 2-input difference (differential) amplifier
> http://en.wikipedia.org/wiki/Differential_amplifier#Other_differential_amplifiers

> What you want is to subtract the output of the VCA from the input to
the VCA, so as the VCA passes more signal, the difference between the
input and output decreases.

That's awesome - very cool. Thanks to Neil for this suggestion.

http://www.sdiy.org/juz/vca_comparison_03.jpg

The response of the 2164 and the single tranny VCA are much closer now - except for the fact that the single Tranny VCA seems to be 'scaled' about 0.5V higher(?) than the 2164.

http://www.sdiy.org/juz/vca_data_01.jpg

I've plotted the two data sets against each other. If I 'scale' (sorry - can't think of a better term than that) the tranny VCA down(?) by 0.5V - so that e.g. the figures for 0.5V are now at 0.0V - you can see the 'error' margin (OK, it's not really error...) is very slim - under the circumstances.

So the next question is - how do I translate this into a change in the VCA? My gut says I need to make the VCA more responsive to current - so e.g. what took 2.00V (0.84V PP) will only take 1.50V to give 0.84V PP... does that make sense?

Any clues on this?

Thanks again - just when I thought I'd dragged every last bit improvement out of this little baby.

Justin