FBN Software proudly presents CHEBFILT. CHEBFILT is an analysis
tool used to find the roots of a Chebycheff or a Butterworth filter.
The program will find the roots of a 1 to 10 pole filter. The roots of
the filter equation are useful in the implementation of active filters
using the state variable method. The program finds the equivalent
lowpass model of the filter and will then transform the filter into a
bandpass, highpass, or notch filter if you so desire. The program also
will provide you with a frequency response of the filter that you can
print out or view on the CRT. CHEBFILT.EXE has been compiled using QB
4.0. QB 4.0 emulates the 8087 output.


The following is a commented operational session with the program
to show you how it works (comments are in parentheses). The text is a
snapshot of the program screens. The filter to be designed is a 4 pole
bandpass Chebycheff filter with a center frequency of 3000 Hz and a
bandwidth of 300 Hz.

***************************************************************************

CHEBYCHEFF AND BUTTERWORTH FILTER GENERATOR



TO SELECT BUTTERWORTH SET RIPPLE EQUAL TO ZERO

By FBN Software

***************************************************************************

For Low Pass Filter Model Input :

NUMBER OF POLES, DB RIPPLE: ? 4,1

3 DB BANDWIDTH : ? 1

(Opening Screen with responses - note a 1 Hz bandwidth was assumed for the
lowpass model, but you could have specified 300 Hz if you had wanted to. If
you were designing a lowpass filter, you can specify the true bandwidth of
the filter.)

DO YOU WANT :

(P)OLE/ZERO PRINTOUT
(F)REQUENCY RESPONSE VIEW - PRINTOUT
(S)TORE RESPONSE IN FFT FORMAT
(N)EW FILTER
(T)RANSFORMATION LP TO HP, NOTCH OR BP
(E)XIT PROGRAM

? t



( options screen - note the output can be saved (S) in FFT format for
use with an FFT program that will be provided later.)



(B)ANDPASS
(H)IGHPASS
(N)OTCH

? b


CENTER FREQUENCY ? 3000

BANDWIDTH? 300

( screen showing the (T) transformation options - b for bandpass and the
center frequency and bandpass selection)


DO YOU WANT :

(P)OLE/ZERO PRINTOUT
(F)REQUENCY RESPONSE VIEW - PRINTOUT
(S)TORE RESPONSE IN FFT FORMAT
(N)EW FILTER
(T)RANSFORMATION LP TO HP, NOTCH OR BP
(E)XIT PROGRAM

? f

(automatic return to options where a F - frequency response was chose)

INPUT FREQUENCY RANGE :

START FREQUENCY? 30
END FREQUENCY ? 300000
Frequency Step Size (-N Gives N equal LOG Steps) ? -41

OUTPUT TO (P)RINTER OR TO (C)RT ? c
Save Plot to file for graphic display (Y/N) ? y
File Name (D:\Path\Name.ext) ? BPFILT.DTA

( The subroutine was entered with a start frequency of 30 Hz and an end
frequency of 300000 Hz. The data is to be presented in logarithmic
frequency steps and the data will be saved to disk)

FREQ. - Hz AMPL. - dB PHASE - Deg. DELAY - Sec.

3.00E+01 -239.92 -0.09 -6.79E-04
3.78E+01 -231.91 -0.11 -6.79E-04
4.75E+01 -223.91 -0.14 -6.79E-04
5.99E+01 -215.91 -0.18 -6.79E-04
7.54E+01 -207.90 -0.22 -6.80E-04
9.49E+01 -199.88 -0.28 -6.81E-04
1.19E+02 -191.86 -0.36 -6.82E-04
1.50E+02 -183.83 -0.45 -6.84E-04
1.89E+02 -175.78 -0.57 -6.87E-04
2.38E+02 -167.70 -0.71 -6.92E-04
3.00E+02 -159.57 -0.90 -6.99E-04
3.78E+02 -151.36 -1.14 -7.12E-04
4.75E+02 -143.03 -1.45 -7.32E-04
5.99E+02 -134.50 -1.86 -7.66E-04
7.54E+02 -125.64 -2.40 -8.23E-04
9.49E+02 -116.24 -3.14 -9.23E-04
1.19E+03 -105.89 -4.23 -1.11E-03
1.50E+03 -93.79 -6.00 -1.52E-03
1.89E+03 -78.11 -9.44 -2.66E-03
2.38E+03 -52.67 -19.82 -8.66E-03

PRESS ENTER TO CONTINUE?

(first display of response showing amplitude, phase and group delay of the
filter.)



DO YOU WANT :

(P)OLE/ZERO PRINTOUT
(F)REQUENCY RESPONSE VIEW - PRINTOUT
(S)TORE RESPONSE IN FFT FORMAT
(N)EW FILTER
(T)RANSFORMATION LP TO HP, NOTCH OR BP
(E)XIT PROGRAM

? p

OUTPUT TO (P)RINTER OR (C)RT ? c

(auto return to options menu where a Pole/Zero printout was selected.)


POLE VALUES

I REAL IMAGINARY

1 -22.11633 3148.365
2 -99.2835 3112.546
3 -91.90927 2881.363
4 -20.02981 2851.34
5 -20.02982 -2851.34
6 -91.90928 -2881.363
7 -99.28349 -3112.546
8 -22.11632 -3148.365


ZERO'S

1 0 0
2 0 0
3 0 0
4 0 0

PRESS ANY KEY TO RETURN TO THE MENU

(filter poles and zero's. Note their are 8 complex pole pairs since it is
a bandpass filter.)

Hints : Backup the programs and data before running them. The program
is now limited to show frequency responses down to -400 dB. Larger
ratio's may cause the machines's mathematical limits of calculations to
be exceeded and cause the program to bomb. A rough estimate of the roll
off of the filter is to assume -20 dB of attenuation for each pole per
decade of frequency from the band edge of the filter. Thus the
approximate attenuation of our example filter should be at least 4*(-20)
or -80 dB at 300 Hz. Note the actual response is approximately -160 dB
due to the high Q (10) of the filter. Limit the frequency spand of your
response until you see it approach -400 dB if you really want to see
how the filter performs " in the mud ".

This filter response has been captured and saved to disk as
BPFILT.DTA. You can use the program FILTF.EXE to view the saved
response of this filter or any other filter that you may design
provided that you have a CGA video output capability.



Problems & Freeware Pitch : I offer this program to the general public as
"Freeware". By this I mean you can make copies of the ARC file and give
them to your friends. Just be sure to include all files (Including
CHEBFILT.DOC) in CHEBFILT.ARC. If you have problems and want support,
I'll provide it to you for a negotiable fee (depending on the
difficulty of the problem). If you want to modify CHEBFILT on your own
(or you're just curious) I'll provide you with the source code for
CHEBFILT (and to PLOTF) on a 360k 51/4 inch disk for a $10.00 fee.

Who am I and where do you send your money :

Send checks (no plastic) to :

William L. Gill
12713 Brunswick Lane
Bowie,Md. 20715

My qualifications : I am an Electrical Engineer with 27 years of
experience. My specialty is Signal Processing. I am a former
applications engineer for Westinghouse Integrated Circuits. I have
approximately 15 years of part time teaching experience (Application of
Digital Integrated Circuits, and the Application of Linear Integrated
Circuits). So I have a strong background in circuit analysis and
performance.

Why am I doing this ? I don't really expect to make money from this
program. It is an engineering program and should therefore have a
limited appeal. I think it provides a tool that some of you will want
to try. The fees that I am imposing are to prevent "nuisance calls".