[sdiy] MOTM 440 4P LPF questions

[Rutger Vlek]

Hi David,

I have been working on my own variation of a discrete OTA lately, and I have been wondering about these things too. After many Spice simulations and some modeling on my Clavia G2, I think I can try to give some answers:

The biggest difference in sound between "discrete OTAs" and "IC-based" OTAs is the way they saturate, and this saturation is the biggest reason for sounding "fat" or "full of character". The saturation has nothing to do with the discreteness (!), but with the typical designs used. The discrete OTA designs I've seen all use fewer transistors than typical IC designs, presumably since it's hard to get the discrete transistors properly matched (or expensive to buy matched ones) and they take more space. Due to these design differences, many discrete OTAs exhibit asymmetric distortion, while most IC OTAs distort in a symmetric way. 

Another factor is of course that most IC designs are hard to come by and getting pretty old. The internal transistors don't have the specs of today's discrete transistors, which could make a difference in terms of noise or saturation curve. The simplicity of the typical discrete OTA design also results in a very simple/short signal path, and as most guitarists (using effects) can tell you: keeping the signal path short gives you the most "direct" sound. Tiny distortions (spectral or phase) easily stack up along a long signal path.

So to conclude, I believe that if an identical OTA design would be implemented in (matched!) discrete components and ICs, using identical transistor technology, they would sound the same.

Best,

Rutger


> 3) Why should a "discrete OTA" give any better results than an IC-based OTA
> or VCA?  After all, an IC is discrete on the inside.  (A direct comparison
> of four different 4P filters, some discrete and some IC-based, on
> Muffwiggler today would seem to debunk this whole "discrete" business, given
> that they all sound identical, which suggests that this whole "discrete"
> argument might just belong in the same category as the various gold-plated
> audiophile claims we often laugh about here.  I'm just curious as to the
> reasons why such circuits would or should be better.)
> 
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at dropmix.xs4all.nl
> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy