[music-dsp] Darlington lattice filter decomposition

Venkatasubramanian V venkatv at ee.iisc.ernet.in
Tue Feb 4 03:57:00 EST 2003

> Does anyone know about decomposition of given xfer function into lattice-scheme elements:
>    IN >--[ Z1 ]-.-[ Z2 ]-.-  ...  [ Zn ]-.---< OUT
>                 !        !               !
>                 -        -               -
>                 Y1       Y2              Yn
>                 -        -               -
>                 !        !               !
>                GND      GND             GND
> For LPFs Z-s are usually the inductors and Y-s is the caps except Yn which is the R. 
> zfer func is 1/(sum(Ai*p^i)). I clearly remember I have read this somewhere somewhen but 
> cant find it now.

I don't remember this well either, but if what you are asking is what I
vaguely remember (Can you get more non-committal than that?) then you must
express the xfer fn as a continued fraction. This answer is by instinct
rather than analysis. (So, one can get more non committal, it seems:)

Venkatasubramanian V

Learning Systems and Multimedia Lab,
Electrical Engineering Department,
IISc, Bangalore - 012

More information about the music-dsp mailing list