Theory for Korg35 (Re: [sdiy] Multi-Input filters)

[Osamu (Sam) HOSHUYAMA]

Hello Paul and List,

Paul Perry wrote:
> 
> At 08:58 AM 12/02/03 +0900, you wrote:
> 
> >In a saturated transistor, normal transistor and inverted transistor
> >(C and E are reversed) work simultaneously.  The CB junction is forward
> >biased, which works as a variable dynamic resistance like the EB
> >junction
> >with ordinary ladder filters.
> >
> 
> Sam, is this how the Korg MS20 VCA works?

Well, this is not sufficient explanation even for the VCF.  We have to 
consider conductance/resistance between C and E.

My understanding is superficial.  The following is my best efforts.
If you need the validity of Ebers-Moll model, I don't know.

In Ebers-Moll model, not only normally operated transistor, the
influence by inverted transistor is also taken into account. 
The influence by the inverted transistor is the most important 
and difficult thing to understand.  I have to use equations to take 
into account the influence.  I'm sorry.

The equations (70) and (71) in the course note indicate that both Ic and 
Ie (the currents at C and E) can be expressed as some exponential-like 
functions of Vc of Vb (Voltages at collector and base.).  (Vcb in the 
equations is equal to (Vc - Vb).)

  
http://www.rpi.edu/~schubert/Courses/ECSE-6290%20SDM-2/1%20BJT-3%202nd%20order%20effects.pdf

As Harry explained, when Vc is sufficiently small, the relationship
between Vc and Ic and Ie can be approximated by linear functions, where
conductance can be defined.  (Ic and Ie are not exactly the same 
usually, but not that far different.)  The conductance (Ic and Ie) 
varies with Vb.  It can be considered as a variable 
conductance/resistance with an exponential-like response.

The difference and the bias of Ic and Ie (mainly the current through 
base, I guess) causes thump.  Of course, distortion is inevitable.

When the base is current driven, not voltage driven.  Eq. (70) and 
(71) have to be rewritten so that Ib is constant.  I've never tried 
it but, its response (relationship between CV and conductance) will 
be more linear.  Quantitative analysis is needed.

Maybe I'm wrong.  Correct explanations in better English would be
appreciated.

Sam Hoshuyama
-- - ---  - -- ------ - -------- -- - ---- -
Osamu (Sam) HOSHUYAMA      houshu at rogers.com
                              Ottawa, CANADA
--------------------------------------------