[sdiy] Analog computers

[Czech Martin]

The 1000 state variable filters were made to filter some
sound, the structure of the filter is known.
The main problem of the OTA (distortion vs. noise)
has not really detrimental effects on the filter,
perhaps this adds sime spice to the sound, but
the overall behaviour is not unexpected (well,
the op amps and otas have some phase lag, which
indeed causes problems, but that can be cured with
lead structures).

But I understood the question as to have a generall
programable analog computer. In this
case the structure you'll build in the end is not
at all clear, only that of the components you have.
I'm not so sure if OTA effects then create artefacts that
fumble the response in some systems you want to compute.

That should be really interesting if you compute a
"chaotic" equation.


m.c.

-----Original Message-----
From: jhaible at debitel.net [mailto:jhaible at debitel.net]
Sent: Mittwoch, 9. Juli 2003 12:50
To: Czech Martin
Cc: tim102 at tstinchcombe.freeserve.co.uk; Tim Ressel; Synth-Diy
Subject: RE: [sdiy] Analog computers



> If you include the integrator resistor, i.e. patch
> behind that resistor, things may get better, but then you need
> a variable resistor, 1000:1 or so. 

A simple OTA will do. It's been done a thousand times
in state variable filters. 

JH.


-------------------------------------------------
debitel.net Webmail