CIC compensator filter specification object
h = fdesign.ciccomp(...,spec,specvalue1,specvalue2,...)
d= fdesign.ciccomp constructs
a CIC compensator specifications object
default values for the properties Fpass, Fstop, Apass, and Astop.
In this syntax, the filter has two sections and the differential delay
fdesign.ciccomp with a design method
dfilt object, a single-rate discrete-time
d= fdesign.ciccomp(d,nsections,rcic) constructs
a CIC compensator specifications object with the filter differential
delay set to
d, the number of sections in the filter
nsections, and the CIC rate change factor
rcic. The default values of these parameters
are: the differential delay equal to 1, the number of sections equal
to 2, and the CIC rate change factor equal to 1.
If the CIC rate change factor is equal to 1, the filter passband response is an inverse sinc that is an approximation to the true inverse passband response of the CIC filter.
If you specify a CIC rate change factor not equal to 1, the filter passband response is an inverse Dirichlet sinc that matches exactly the true inverse passband response of the CIC filter.
d= fdesign.ciccomp(..., constructs
a CIC Compensator specifications object and sets its
spec. Entries in the spec string represent
various filter response features, such as the filter order, that govern
the filter design. Valid entries for
shown in the list below. The strings are not case sensitive.
The string entries are defined as follows:
ap — amount of ripple allowed
in the pass band in decibels (the default units). Also called Apass.
ast — attenuation in the
stop band in decibels (the default units). Also called Astop.
fc — cutoff frequency for
the point 6 dB point below the passband value. Specified in normalized
fp — frequency at the end
of the pass band. Specified in normalized frequency units. Also called
fst — frequency at the start
of the stop band. Specified in normalized frequency units. Also called
n — filter order.
In graphic form, the filter specifications look like this:
Regions between specification values like
transition regions where the filter response is not explicitly defined.
The filter design methods that apply to a CIC compensator specifications
object change depending on the
designmethods to determine
which design method applies to an object and its specification string.
h = fdesign.ciccomp(...,spec,specvalue1,specvalue2,...) constructs
an object and sets the specifications in the order they are specified
spec input when you construct the object.
Typically, when they develop filters, designers want flat passbands and transition regions that are as narrow as possible. CIC filters present a (sinx/x) profile in the passband and relatively wide transitions.
To compensate for this fall off in the passband, and to try
to reduce the width of the transition region, you can use a CIC compensator
filter that demonstrates an (x/sinx)
profile in the passband.
fdesign.ciccomp is specifically
tailored to designing CIC compensators.
Here is a plot of a CIC filter and a compensator for that filter. The example that produces these filters follows the plot.
Given a CIC filter, how do you design a compensator for that
filter? CIC compensators share three defining properties with the
CIC filter — differential delay,
numberofsections; and the usable passband
By taking the number of sections, passband, and differential delay from your CIC filter and using them in the definition of the CIC compensator, the resulting compensator filter effectively corrects for the passband droop of the CIC filter, and narrows the transition region.
As a demonstration of this concept, this example creates a CIC decimator and its compensator.
fs = 96e3; % Input sampling frequency. fpass = 4e3; % Frequency band of interest. m = 6; % Decimation factor. hcic = design(fdesign.decimator(m,'cic',1,fpass,60,fs)); hd = cascade(dfilt.scalar(1/gain(hcic)),hcic); hd(2) = design(fdesign.ciccomp(hcic.differentialdelay, ... hcic.numberofsections,fpass,4.5e3,.1,60,fs/m)); fvtool(hd(1),hd(2),... cascade(hd(1),hd(2)),'Fs',[96e3 96e3/m 96e3])
You see the results in the preceding plot.
Designed to compensate for the rolloff inherent in CIC filters,
CIC compensators can improve the performance of your CIC design. This
example designs a compensator
d with five sections
and a differential delay equal to one. The plot displayed after the
code shows the increasing gain in the passband that is characteristic
of CIC compensators, to overcome the droop in the CIC filter passband.
Ideally, cascading the CIC compensator with the CIC filter results
in a lowpass filter with flat passband response and narrow transition
h = fdesign.ciccomp; set(h, 'NumberOfSections', 5, 'DifferentialDelay', 1); hd = equiripple(h); fvtool(hd);
This compensator would work for a decimator or interpolator that had differential delay of 1 and 5 sections.