SV: Re: [sdiy] CEM3328 seems to only have buffers at the end

[Antti Huovilainen]

On Wed, 22 Feb 2006 jhaible at debitel.net wrote:

> I think now I understand what you mean.
> "when other things are equivalent" is the important point here.

Of course. I should have probably mentioned it earlier, but kinda assumed 
it was a given. Probably because in digital (I'm still too lazy to do much 
analog building) the resonance vs core level can be adjusted with a single 
number and is thus obvious on first glance.

Taking a look at some schematics, it seems that often OTA core levels are 
similar or even lower than Moog ladder levels (comparing say Moog Taurus 
and Roland SH-101), so it seems safe to say that my claims often hold in 
practise too (although not always).

> But you have so many degrees of freedom to design these filters.
> Especially when you're using OTA filters with buffers and resistor
> dividers, you can set your "working level" (and the clipping which
> would occur in another OTA for feedback) just as you like.

Nitpicking, but the same goes for Moog. Input divider sets core level, 
output differential amplifier the resonance distortion (usually pretty 
low when opamp is used for this).

> And with the buffers clipping themselves, unsymmetrically as
> in the 2040 even, you have that wonderful interaction of frequency
> depended differential pair distortion (around the cutoff frequency,
> as you explained), and the frequency independend, but somehow
> transient depended, distortion from the buffers.

I _think_ (read: I've done a few spice simulations but nothing conclusive) 
what happens here is that the base-collector diode becomes active when 
capacitor voltage goes below diff-pair - input. This is somewhat similar 
to tube amplifiers where the grid also acts as a diode thus varying 
the DC voltage over the interstage DC blocking capacitor.

Antti

"No boom today. Boom tomorrow. There's always a boom tomorrow"
   -- Lt. Cmdr. Ivanova