AC-Coupled module inputs

[Don Tillman]

   From: "List, Christopher" <Chris.List at sc.siemens.com>
   Date: Mon, 26 Apr 1999 10:39:12 -0400

   I was pondering something last night about module design
   philosophy. Why would you put ac-coupling caps on module inputs, or
   have them be switchable?  Wouldn't it make more sense to have all
   inputs uncoupled, and just make sure that any module that is
   supposed to produce or process audio ac signals doesn't have a dc
   offset on it's output?

Yes, the latter apporach is better.

AC-coupled inputs are only useful for the cases where you have no
assurance that your audio signals have no DC offsets.  

The issue gets thorny when your VCOs perform as both audio and low
frequency cv sources.  For audio use you want zero dc offset.  For low
frequency cv use you want your waveforms to start at zero volts and go
from there.  While there's no difference between the two for sine and
triangle waves, the differences are important for sawtooth, square and
PWM waves.

There's a school of synth design that says VCOs should be universal,
we don't want to impose preconceptions or restrictions on their use,
they cost enough as it is and they should be able to be used for
either application.  Like on the ARP 2600.

I'll claim that approach is misguided and you really do want separate
audio and low frequency VCOs.  (Heck, lots of different types of
VCOs!)  And with separate classes of VCOs you can have dc-coupled
inputs throughout.

(This same issue also applies to waveshapers and ring modulators I
guess.)

  -- Don