Arrays and quads

Grant Richter grichter at
Fri Mar 28 07:36:31 CET 1997

>Arrays ---   
>I've been thinking about the exp. current sorce thing some more...
>There seem to be quite a number of transistor arrays as Rick pointed out 
>in an old post. (Subject: Re: Substrate temperature stabilized;  Date: 
>Thu, 08 Jun 1995).  There is a CA3096 N-P-N/P-N-P Transistor Array.  How 
>many transistors are there and how are they arranged?  I thought that 
>maybe if the PNP and NPN parts are separate (and had matched PNP-PNP and 
>NPN-NPN pairs) one could build a dual Electronotes option 1 - option 2 
>module.  It's probably too much to hope for...

I've built several VCO's using the CA3096. It contains two PNP's and three NPN's.
Just the right number to implement the substrate temperature control circuit from
National's AN-299 and still have the two PNP transistors left to implement
Bernie Hutchins classic exponential converter. This doesn't need the Q81 tempco
then of course and the extra parts needed are cheaper than the tempco. According
to Electronotes the thermally stabilized substrate is superior to the tempco design
because the tempco only gives partial compensation for temperature(3500 ppm vs. 3300? ppm)
whereas the thermal stabilization eliminates the problem altogether.

The design is also used in the Serge NTO which uses a thermally stabilized 
CA3046 for the exponential convertor. Quite rock solid. Other synths using the
CA3046 design are the Norlin Moogs - Prodigy and Source I know for sure
and probably the Micro and Multi (not the Mini which used tempcos).

Be very careful of PCB layout and power wiring. The 70 ma drawn by the heater is enough
to cause voltage drops in traces and voltage shifts in 500ma supplys. Also don't 
substitute the single 741 used in the circuit. I used 1/4 of a LF347 and had to up the
loop cap from .01 to .1 for stability and move it outside to eliminate bias shifts.

Comparing all the VCO's for short term stability (phase wobble) the winner is....
The PAIA experimenters board using a CEM3340 IC and no fancy components.
Probably because the main source of short term pitch instability is component noise
and is not thermally related at all. Long term THERMAL stability only comes into
play at those outdoor festivals where you soundcheck at noon and play at eight on
a windy day.

More information about the Synth-diy mailing list