AW: Secret CEM chip info...
Haible Juergen
Juergen.Haible at nbgm.siemens.de
Wed Feb 19 12:46:00 CET 1997
> I asked Doug about the SSM filter chip versus the CEM. They are in fact
> completely different topologies. The SSM is really a "cheesy" part that has
> 4 CA3080 OTAs and a rather crude exponential generator.
As far as I know, the SSM (2040 that is) OTAs are even more simple than
3080's (two current mirrors omitted), which is imortant for its unique
overdrive
charactersistic.
I don't know anything about the exponential generator. Dave Rossum said
it was a very special design that avoids the need of an external
compensation
capacitor, but he wouldn't go into detail. Can you tell me more details?
> It generates quite
> a bit of third and fifth harmonics, which explains the "fatness".
To some extend, at least. That's what a 3080-based VCF would do also.
Things become much more interesting if you overdrive the 2040 so hard
that the *output* distortion comes into play. (i.e. if you use it
outside
the +/-1V signal range which was suggested in the data sheet.)
> Sometimes
> electrically "crummy" is "superb" mucically. [...]. Sheesh..... The CEM
filters were d esigned to be "cleaner" from a THD+noise value
Very well put. Please, if you have any chance to do it, resurrect the
SSM2040.
Building a *clean* filter may have been difficult back then, but today
it's much more
difficult to reproduce the "character" of a filter like the 2040.
JH.
More information about the Synth-diy
mailing list