How the heck does the MS-50 filter WORK?

[<<marjan>>]

Sean Costello wrote:

> 
> Thanks!  It is starting to make sense now.  Question:  What were the
> dates of these two emails?  Martin referred to an earlier post of his on
> diode ladders, which I would love to see.  If I had a date, I could
> download the whole month's mail and search through it.  (If only
> Synth-DIY had up-to-date archives!)

Is there some archive with 97. ? I know only the one (old ) on
Rick Jensen's  page.

About those posts
 all are 29 Aug 1997

This is Martin's mail you asked :


start quote

Subject: 
        diode ladders
   Date: 
        Fri, 29 Aug 1997 13:28:08 +0200 (MET DST)
   From: 
        Martin Czech <martin.czech at itt-sc.de>
     To: 
        synth-diy at horus.sara.nl



How do ladder circuits work ?

The bias current through a nonlinear device changes the
small signal resistance i.e. a current controlled
resistance.

Ideal diode (and emitter current of transistor as well):

I=Is*[exp(u/Ut)-1] => U=Ut*log(I/Is+1) 

R(I)=dU/dI=Ut/(I+Is) for I>>Is : R(I)=Ut/I

The small signal resistance via the diode or transistor in the
ladder circuit is inverse proportional to the bias current.
Is this correct?

So it is difficult to make a linear response !?

If we chose I=k*exp(a*Uc) (exponential control of bias
current via control voltage Uc), we find:

R(Uc)=Ut/k/exp(a*Uc)=1/k*Ut*exp(-a*Uc)

Now the small signal resistance via the diode or transistor in the
ladder circuit is proportional to some exponential function
of Uc, that's what we need. Add some capacitor and we have first
order lowpass section with exponential control of "R", that is
frequency.
For better offset suppression etc. the whole ladder is made symetrical,
that is : each stage has two resistive devices and one cap between.

O.K.?

But how can a highpass be achieved? Simple exchange of "R" and C won't
work ? Has anybody a schematic of a cascade highpass circuit?

end quote   



marjan