Analog Computer

[Martin Czech]

> Is there one, who has tried an (simple) analog computer (AC) to process
> control voltages or even VCO-signals ?
> After all, the state variable filter is a "output" of the AC-technique.
> I think, this could be interesting, but before building up an AC (even a
> simple one),
> it wanted to get some information.
> 
> By the way : there is a site with very simple add/substr/mult/div,
> integration a.s.o. 
> schematics, I have to take a look, if I find the URL once more. I will
> announce it,
> if I know the URL.
> 
I havn't yet, but I want to.
If you take care that your (new designed) modules accept dc as well as
ac signals, and are able to process dc without offsets etc. you have
pretty much of the stuff you need. You have 2 and 4 quadrant multipliers
(well, disguised as vca and ring mod.).  You have nonlinear functions
(e.g. sine shaper, or other shapers, or J.H.'s interpolator for PWL). You
may add some constant voltage module (10 turn precision pots for fixed
parameters). You have adders (well mixers, switchable polarity for each
input would be good in this case) What's really missing now is a voltage
controlled integrator (we usually call this lp filter). I toyed arround
with a integrator/ssm2018 to have a low offset, low fedthrough integrator,
it worked quite good (the 2018 has voltage output, better use 2218 with
current output), and the time constant is voltage controlled over a wide
scale. There you are.

This is more then the analog computers had that I used on university in
87 or so.  (Yes, they are oldfashioned now, but you learn a lot about
differential equations etc. with these old machines.  Did you know
that I had to use punched cards for laboratory reservation and course
inscription. Why not? At least as durable as a floppy, no format problems,
and enough space for your name, date of birth and id-number).

m.c.