[sdiy] Discrete OTA

[David G Dixon]

> >Not quite following you here.  The dates are ten years 
> apart.  Clearly 
> >Romans invention!
> >
> Crikey, I only looked at the last digit in the year, I didn't 
> expect such a large offset between the two, especially when 
> it was the same month and almost the same day... Well that 
> removes all ambiguity then.

The scary thing is, I believe that, after all this time, I'm still the only
one who has worked this into an actual product.  The Intellijel VCOs use
this scheme, and I don't believe that any other commercially available VCOs
do.  I don't know why, as the Dixie and Rubicon have the reputation of being
some of the most accurate and stable VCOs around, and it is really so much
more convenient than tempco resistors (with their convoluted layouts and
goop) or heaters.

By the way, I came to the conclusion, after much calculation and
experimentation, that the optimum tempco voltage is 274 mV.  Achieving this
is a little tricky because of the 5k input impedance of the 2164 VC pin.
The fact is that the actual gain ratio of the 2164, which is advertised to
be 33 mV/dB or 200 mV/decade (and modeled that way by most simulator
jockeys), is actually between 180 and 190 mV/decade, about 185 mV/decade on
average (IIRC), and this changes the tempco voltage a bit.  There are lots
of such surprises in store for those of us who try to use the 2164 in every
possible way, but it's still a brilliant chip, warts and all!