[sdiy] SSM2040 filter question

[René Schmitz]

Hi Antti et al.,

> I would like someone explain how SSM2040 differs from 4 OTAs. To me each
> section looks identical to CA3080 + darlington follower.

The difference is that the CA3080 can sink and source current to both 
rails, while the 2040 gain cell can sink only to -800mV below "GND" 
(which is 5V in my version). As I explained earlier, this leads to a 
dynamic nonlinearity.

> Also, how is the feedback amplitude limited since there is no nonlinearity
> in the feedback path in that version?

But, there is one inside the loop and thats all it takes to delimit the 
amplitude.

> There are then three possible conclusions:
> - Input goes close to supply rails and thus there is asymmetric distortion
>   (unlikely to sound good there).
> - Darlington pair has highish output impedance, thus distorting the audio.
> - SSM2040 doesn't clip asymmetrically and sounds good for other reasons.

I'd say, get your self a Spice simulator, and see what happens for 
yourself.

Cheers,
  René


-- 
uzs159 at uni-bonn.de
http://www.uni-bonn.de/~uzs159