Bessel analog filter design
returns the transfer function coefficients of an
a] = besself(
lowpass analog Bessel filter, where
Wo is the angular
frequency up to which the filter's group delay is approximately constant. Larger
n produce a group delay that better approximates
a constant up to
function does not support the design of digital Bessel filters.
Design a fifth-order analog lowpass Bessel filter with approximately constant group delay up to rad/second. Plot the magnitude and phase responses of the filter using
[b,a] = besself(5,10000); freqs(b,a)
Compute the group delay response of the filter as the derivative of the unwrapped phase response. Plot the group delay to verify that it is approximately constant up to the cutoff frequency.
[h,w] = freqs(b,a,1000); grpdel = diff(unwrap(angle(h)))./diff(w); clf semilogx(w(2:end),grpdel) xlabel('Frequency (rad/s)') ylabel('Group delay (s)')
Design a 12th-order bandpass Bessel filter with the passband ranging from 300 rad/s to 500 rad/s. Compute the frequency response of the filter.
[b,a] = besself(6,[300 500],'bandpass'); [h,w] = freqs(b,a);
Plot the magnitude and phase responses of the filter. Unwrap the phase response to avoid and jumps and convert it from radians to degrees. As expected, the phase response is close to linear over the passband.
subplot(2,1,1) plot(w,20*log10(abs(h))) ylabel('Magnitude') subplot(2,1,2) plot(w,180*unwrap(angle(h))/pi) ylabel('Phase (degrees)') xlabel('Frequency (rad/s)')
Wo— Cutoff frequency
Cutoff frequency, specified as a scalar or a two-element vector. A cutoff frequency is an upper or lower bound of the frequency range in which the filter's group delay is approximately constant. Cutoff frequencies must be expressed in radians per second and can take on any positive value.
Wo is scalar, then
besself designs a lowpass or highpass
filter with cutoff frequency
Wo is a two-element vector
[w1 w2], where
besself designs a bandpass or bandstop
filter with lower cutoff frequency
higher cutoff frequency
ftype— Filter type
Filter type, specified as:
'low' — a lowpass filter with cutoff
the default for scalar
'high' — a highpass filter with cutoff
'bandpass' — a bandpass filter of order
Wo is a
'bandpass' is the default
Wo has two elements.
'stop' — a bandstop filter of order
Wo is a
a— Transfer function coefficients
Transfer function coefficients of the filter, returned as row vectors of
n + 1 for lowpass and highpass filters
n + 1 for bandpass and bandstop filters.
The transfer function is expressed in terms of
k— Zeros, poles, and gain
Zeros, poles, and gain of the filter, returned as two column vectors of
n for bandpass and
bandstop designs) and a scalar. The transfer function is expressed in terms
D— State-space matrices
State-space representation of the filter, returned as matrices. If
n for lowpass and
highpass designs and
m = 2
n for bandpass
and bandstop filters, then
m × m,
B is m × 1,
C is 1 × m, and
D is 1 × 1.
The state-space matrices relate the state vector x, the input u, and the output y through
besself designs analog Bessel filters, which are characterized by
an almost constant group delay across the entire passband, thus preserving the wave
shape of filtered signals in the passband.
Lowpass Bessel filters have a monotonically decreasing magnitude response, as do lowpass Butterworth filters. Compared to the Butterworth, Chebyshev, and elliptic filters, the Bessel filter has the slowest rolloff and requires the highest order to meet an attenuation specification.
For high-order filters, the state-space form is the most numerically accurate, followed by the zero-pole-gain form. The transfer function coefficient form is the least accurate; numerical problems can arise for filter orders as low as 15.
besself uses a four-step algorithm:
Find lowpass analog prototype poles, zeros, and gain using the
Convert the poles, zeros, and gain into state-space form.
If required, use a state-space transformation to convert the lowpass filter into a bandpass, highpass, or bandstop filter with the desired frequency constraints.
Convert the state-space filter back to transfer function or zero-pole-gain form, as required.
 Parks, Thomas W., and C. Sidney Burrus. Digital Filter Design. New York: John Wiley & Sons, 1987.