Digital Signal Processing Reference
In-Depth Information
Substituting
z ¼ e
j
U
, we get the residue as
R ¼ Ve
jf
x
Hðe
j
U
Þ
YðzÞ
z
¼
Ve
jf
x
Hðe
j
U
Þ
z e
j
U
þ
sum of the rest of partial fractions
(D.6)
and multiplying z on both sides of Equation
(D.6)
leads to
YðzÞ¼
Ve
jf
x
Hðe
j
U
Þz
z e
j
U
þ z
sum of the rest of partial fractions
(D.7)
Taking the inverse z-transform leads to two parts of the solution:
yðnÞ¼Ve
jf
x
Hðe
j
U
Þe
j
U
n
þ Z
1
ð
z
sum of the rest of partial fractions
Þ
(D.8)
(D.9)
and the transient response
y
tr
ðnÞ¼Z
1
ð
z
sum of the rest of partial fractions
Þ
(D.10)
Note that since the digital filter is a stable system, and the locations of its poles must be inside the unit
circle on the z-plane, the transient response will settle to zero eventually. To develop the filter
magnitude and phase responses, we write the digital steady-state response as
y
ss
ðnÞ¼V
Hðe
j
U
Þ
e
j
U
þjf
x
þ
:
Hðe
j
U
Þ
(D.11)
Amplitude of the steady-state output
Amplitude of the sinusoidal input
Magnitude response
¼
(D.12)
Hðe
j
U
Þ
¼
Hðe
j
U
Þ
¼
V
V
¼
e
jf
x
þj
:
Hðe
j
U
Þ
e
jf
x
¼ e
j
:
Hðe
j
U
Þ
¼
:
Hðe
j
U
Þ
Phase response
(D.13)
Thus we conclude that
Frequency response:
Hðe
j
U
Þ¼HðzÞj
z¼e
j
U
(D.14)
Since
Hðe
j
U
Þ¼jHðe
j
U
Þj
:
Hðe
j
U
Þ
Magnitude response:
Hðe
j
U
Þ
(D.15)
Phase response:
:
Hðe
j
U
Þ
(D.16)
Search WWH ::
Custom Search