AW: VCO temp compensation ideas

[Rich Nelson]

The Dark force of dance wrote:
> 

> Anyway I'm sure that using thermistors use to be the common way of
> stabilizing VCOs. I can't remember exactly but you pick a positive or
> negative temperature coefficient depending on your drift problem and you
> trim it out so that it's effect on the circuit matches it's drift.
> Presumably you could probably even whack it across the control voltage and
> adjust that up or down to suit.
> 

You don't pick depending on your drift problem.  Exponential converters
always drift the same direction, as dictacted by the formula for the
exponential function, which I would give if it were not such a mess to
write on a word processor!

And what you __are__ doing is compensating the control voltage, since
the drift we are concerned with is the drift in the CV exponential
converter.

At around room temp., the drift is about .33% / degree C.  So you need
to use a thermistor which has about the same change in R/temp.  The gud
ol' Q31 has a temp. coef. of 3500ppm/deg C, so it is pretty close.

But you CAN use almost any thermistor with a fairly linear region around
room temp, either positive or negative temp coef. as long it is 3300ppm
or greater.  You simply put it in series with a resistor, so that the
change in resistance of the pair is 3300ppm/deg C  (keeping in mind that
depending on the type, resistors will themselves have a positive tempco
of between about 100ppm and 500ppm).

You can use negative coef. by changing its position in the circuit.  For
instance, if you place the resistor/thermistor voltage divider
combination on the output of your CV summer, with a positive tempco, you
place it in the lower leg, but with a negative tempco, you place it in
the upper leg.

rich
<<pantechnicon>>