diode capacitance (was:Re: [sdiy] Doepfer Schematics)

[Magnus Danielson]

From: jhaible at t-online.de (jh.)
Subject: diode capacitance (was:Re: [sdiy] Doepfer Schematics)
Date: Mon, 7 May 2001 21:42:06 +0200

> > [Czech Martin]  I may be wrong, but did anybody consider diffusion
> > capacitance of the ladder junctions? I mean they are forward biased
> > and show as a complex impedeance. Both, real and imag part depend
> > on frequency. I did no simulations or calculations, but perhaps a
> > few additional zeros and poles will show up, maybe at a too high
> > frequency...
> 
> This is interesting. Must check if these caps just introduce an extra
> attenuation (with the shunt caps of the ladder), or if there are zeroes
> introduced. Most probably out of the "interesting" range.

First, I wounder if this capacitance makes any major contribution at
the frequencies we are discussing...

Then, one must recall that the capacitance over the NP junction is
actually a normal cap with a voltage sensitive cap in parallell.
Then consider the transconductance over the NP junction as a variable
resistor...

If we would model the transistors and caps as an R-C bridge, then the
modified model of the transistor would make this into an RC-C bridge.
This would add zeros and move the poles. In the loopbacked case poles
would behave basically the same but the filter effect would without
the extra capacitance... but the behaviour will be diffrent never the
less.

But then... given the static bias voltage of these, then there would
be static zeros and movement of poles. The changes of voltage will
modulate these properties, but in a most nonlinear fashion... but
then, the voltage sensitive caps will most probably only contribute
some where as the operating points of the NP-junctions on the
transconductance will most probably do the most of it...

Hmm... when they rolled the first semiconductors there at Bell
Labs... why couldn't they have done it properly? ;)

Cheers,
Magnus