Compensating multi-stage feedback (was: RE: all tranny vca+ )

[Martin Czech]

we are getting religious ... What's the problem with opamps?

They where once intended for precision DC amplification and low frequency
calculation.

The problem seems to me that people use the wrong type of opamp in the
wrong place and in the wrong way.

If a circuit designer uses a 741 as audio device, it is his fault,
not the fault of the opamp designer if nasty audio artefacts are audible.

If the input stage (or the millering stage) is pretty fast, and the
thing remains in linear operation there is no problem. Problems by large
transients could be circumvented by limiting the input bandwidth to some
reasonable value (for me it is something like 20kHz) with a passive network.
If we aggree that this will not compromise our music, we'll notice that bad
input transients are now harmless, because thelinear input "guard" network slows 
them down to something reasonable. Even the unfamous 741 can deal with signals
that way, but in this case the bandwidth has to be degraded.

If you take a look to MOSFET power amps, you'll sometimetimes notice opamp
like structures, a very fast input differential stage, a very fast cascode
stage, and the dominant pole sits really in the power stage, ie. Cgs and Cgd
of the mosfets. It it very likely that the inout stage will never fall out of
linear mode this way, the rest is always done with an input "guard" filter.

m.c.