[sdiy] "Substrate Oven" in VCO and VCF again

[Nils Pipenbrinck]

Hi folks.

We had the heated expo-pair topic recently, and since I still have some 
CA3046 at hand I thought to give it a try.

First I checked a bunch of schematics and I am confused: Pretty much all 
of them violate the maximum specs of the CA3046 in one way or another.

Here is what I pulled from the CA3046 data-sheet:

  - Max collector current: 50mA
  - Max power dissipation per tranny: 300mW
  - Max collector to substrate voltage: 20V


Now let's start with the classic AN-299:

  http://users.ece.gatech.edu/lanterma/sdiy/datasheets/opamp/AN-299.pdf

Here, at an opamp output voltage of 10V I get (by simulation):

   Base Voltage: 6.6V
   Heater Collector current: 160mA (3 times as much as allowed)
   Heater Power dissipation around 1.6W (5 times as much as allowed)


Doepfer A-110:

   http://www.doepfer.de/a110_s.gif

There is no voltage divider in front of the base of the heater 
transistor, but the high base resistor somewhat limits the current. This 
is of course dependent on the hFE of the transistor. With simulation I 
still see the transistor dissipating about 1W and the current is out of 
spec as well.


And then there is Moog Prodigy:

 
http://www.emusic-diy.org/MoogManuals/Prodigy?action=AttachFile&do=get&target=prod9.gif

I haven't estimated the power dissipation or collector current (will do 
so later), but the way Moog connects the substrate would violate the 
maximum collector to substrate voltage (24 instead of 20V).

Only good old Electronotes S-019 seems to be fine, albeit is is a bit on 
the weak side with just about 200mW dissipation.

   http://electronotes.netfirms.com/s019.pdf


So with all these maximum spec violations these circuits still work fine 
in practice. Have I done something stupid while calculating the 
collector currents and dissipation or are the transistors really 
operated that much out of their spec?

Best,
   Nils