[sdiy] Analog Computers, Rocket City

[Donald Tillman]

On Jan 26, 2013, at 9:02 PM, "Lanterman, Aaron" <lanterma at ece.gatech.edu> wrote:

> The EML 101 design has an analog computer-ish feel to it. The VCF has a weird integrator structure that uses two capacitors per integrator. Someone -- Harry Bissell, maybe? I can't remember now -- mentioned that was a design that showed up in old-school analog computers.


I'm intrigued... here's the schematic:

    http://www.synfo.nl/servicemanuals/EML/EML-101_SCHEMATICS.pdf

I've always known the differential integrator as a classic opamp circuit, 'never heard about it used in analog computers before.

So in this circuit they're taking the balanced outputs of the transconductance pairs and sending them to differential integrator inputs.  That seems elegant.

But not really.

The input impedances of the differential integrator are wildly different, and the positive input voltage is seen on the negative input.

(And this is probably why I haven't seen it used in analog computing.  You'd have to buffer each input first.)

So in this case the differential integrators will place some weird loads on the transconductance pairs, probably creating some unique nonlinearities.  But hey, maybe it sounds good.

  -- Don

--
Don Tillman
Palo Alto, California
don at till.com
http://www.till.com