Digital Signal Processing Reference
In-Depth Information
Example 6.4 Butterworth Step Invariant Filter Design
Problem:
Determine the step invariant digital filter for a second-order
normalized Butterworth approximation function, as shown below. Determine the
differences that result for sampling periods of
T
= 1.0 sec and
T
= 0.1 sec.
1
H
(
s
)
=
2
s
+
1
4142
⋅
s
+
1
Solution:
Again, we first determine the output of the analog system to an
input step function and then use partial fraction expansion to determine the
individual terms.
1
0
+
j
0
0
−
j
0
G
(
s
)
=
−
−
s
s
+
0
7071
+
j
0
7071
s
+
0
7071
−
j
0
7071
The step response can then be determined by finding the inverse Laplace
transform of
G
(
s
), as indicated below:
−
(
0
.
7071
+
j
0
.
7071
)
t
−
(
0
.
7071
−
j
0
.
7071
)
t
g
(
t
)
=
1
.
0
−
(
0
.
5
+
j
0
.
5
⋅
e
−
(
0
.
5
−
j
0
.
5
⋅
e
Then, after sampling at intervals of
T
, the discrete-time step response is
determined to be
−
(
0
.
7071
+
j
0
.
7071
)
nT
−
(
0
.
7071
−
j
0
.
7071
)
nT
g
(
nT
)
=
1
.
0
−
(
0
.
5
+
j
0
.
5
⋅
e
−
(
0
.
5
−
j
0
.
5
⋅
e
The discrete-time response to the step input can then be determined by using
the
z
-transform.
1
0
+
j
0
0
−
j
0
G
(
z
)
=
−
−
−
1
(
0
7071
+
j
0
7071
)
T
−
1
(
0
7071
−
j
0
7071
)
T
−
1
1
−
z
1
−
e
z
1
−
e
z
Now, by combining these terms over a common denominator (and performing
a considerable amount of complex algebra), we have the system response to a step
input:
−
1
−
0
707
T
−
1
z
+
e
[(sin(
0
707
⋅
T
)
−
cos(
0
707
⋅
T
)]
⋅
z
G
(
z
)
=
−
1
−
0
707
⋅
T
−
1
−
1
414
⋅
T
−
2
(
−
z
)
⋅
[
−
2
⋅
e
cos(
0
707
T
)
⋅
z
+
e
z
]
−
1
414
T
−
2
−
0
707
T
−
2
e
z
−
e
[sin(
0
707
⋅
T
)
+
cos(
0
707
⋅
T
)]
⋅
z
+
−
1
−
0
707
⋅
T
−
1
−
1
414
⋅
T
−
2
(
−
z
)
⋅
[
−
2
⋅
e
cos(
0
707
⋅
T
)
⋅
z
+
e
z
]