AW: Temperature control for transistors/

[Haible_Juergen#Tel2743]

> You could use this circuit to temperature-stabilize dual transistors
> in exp-generators. By far the best way to do this is substrate
> heating, i.e. heat the silicon chip the transistors are on directly.
> You need at least two extra transistors on the wafer to do this,
> one transistor as the heater, the other as the sensor.

With a clever design, you only need *one* additional transistor!
I think I have seen it in the Micromoog: They use the extra transistor
as a heater, but they measure the BE voltage of one transistor from the
exponential converter, to read the temperature!
This should work fine as long as the current tru this "sensor" transistor
is constant. So we could use the transistor in the "reference current"
path of the expo converter, as long as we do not modultate the reference
current for linear FM.
The 3046/3086 surely not being ideal devices, I found that the tracking of
its two halves is excellent in my SEM-VCO clone, so building a Dual VCO
with one 3086, where the fifth transistor is a heater could be a good 
solution!
(But I haven't trie it exactly; I used the tempco method for my SEM-VCOs,
and in the Micromogg it's a VCO and a VCF that track together, but with
the VCF being used as 2nd VCO, and high effords to make it track ...)

JH.