Ee101 Bode 1 PDF

EE101: Bode plots
M. B. Patil
mbpatil@ee.iitb.ac.in
Department of Electrical Engineering
Indian Institute of Technology Bombay
M. B. Patil, IIT Bombay
What is deciBel (dB)?
* The unit dB is used to represent quantities on a logarithmic scale.

* Because of the log scale, dB is convenient for representing numbers that vary in
a wide range.

a wide range.
* log scaling roughly corresponds to human perception of sound and light.

a wide range.
* log scale allows and to be replaced by + and simpler!

a wide range.
* The unit Bel was developed in the 1920s by Bell Labs engineers to quantify
attenuation of an audio signal over one mile of cable.

a wide range.
Interesting facts:
- Alexander Graham Bell, who invented the telephone in 1876, could never talk to his
wife on the phone (she was deaf).

a wide range.
Interesting facts:
- Bell considered the telephone an intrusion and refused to put one in his office.

a wide range.
Interesting facts:
- Bell considered the telephone an intrusion and refused to put one in his office.
* Bel turned out to be too large in practice deciBel (i.e., one tenth of a Bel).
* dB is a unit that describes a quantity, on a log scale, with respect to a reference

quantity.
X (in dB) = 10 log10 (X /Xref ).

quantity.
For example, if P1 = 20 W and Pref = 1 W ,
P1 = 10 log (20 W /1 W ) = 10 log (20) = 13 dB.

quantity.
P1 = 10 log (20 W /1 W ) = 10 log (20) = 13 dB.
* For voltages or currents, the ratio of squares is taken (since P V 2 or P I 2
for a resistor).

quantity.
P1 = 10 log (20 W /1 W ) = 10 log (20) = 13 dB.
for a resistor).
For example, if V1 = 1.2 V , Vref = 1 mV , then
`
V1 = 10 log (1.2 V /1 mV )2 = 20 log 1.2/103 = 61.6 dBm.

quantity.
P1 = 10 log (20 W /1 W ) = 10 log (20) = 13 dB.
for a resistor).
For example, if V1 = 1.2 V , Vref = 1 mV , then
`
V1 = 10 log (1.2 V /1 mV )2 = 20 log 1.2/103 = 61.6 dBm.

* The voltage gain of an amplifier is
AV in dB = 20 log (Vo /Vi ),
with Vi serving as the reference voltage.
Example
Vi
Amplifier
Vo
Given Vi = 2.5 mV and AV = 36.3 dB,

compute Vo in dBm and in mV.
(Vi and Vo are peak input and peak output voltages, respectively).
Example
Vi
Vo
Amplifier

Method 1:
Vi = 20 log
2.5 mV
1 mV
= 7.96 dBm
Example
Vi
Vo
Amplifier

Method 1:
Vi = 20 log
2.5 mV
1 mV
= 7.96 dBm
Vo = 7.96 + 36.3 = 44.22 dBm
Example
Vi
Vo
Amplifier

Method 1:
Vi = 20 log
2.5 mV
1 mV
= 7.96 dBm
Vo = 7.96 + 36.3 = 44.22 dBm
Vo (in mV )
Since Vo (dBm) = 20 log
,
1 mV
Vo = 10x 1 mV , where
x=
1
Vo (in dBm)
20
Example
Vi
Vo
Amplifier

Method 1:
Vi = 20 log
2.5 mV
1 mV
= 7.96 dBm
Vo = 7.96 + 36.3 = 44.22 dBm
Vo (in mV )
,
1 mV
x=
1
Vo (in dBm)
20
Vo = 162.5 mV .
Example
Vi
Vo
Amplifier

Method 1:
Method 2:
Vi = 20 log
2.5 mV
1 mV
= 7.96 dBm
Vo = 7.96 + 36.3 = 44.22 dBm
Vo (in mV )
,
1 mV
x=
1
Vo (in dBm)
20
Vo = 162.5 mV .
AV = 36.3 dB
20 log AV = 36.3 AV = 65.
Example
Vi
Vo
Amplifier

Method 1:
Method 2:
Vi = 20 log
2.5 mV
1 mV
= 7.96 dBm
Vo = 7.96 + 36.3 = 44.22 dBm
Vo (in mV )
,
1 mV
x=
1
Vo (in dBm)
20
Vo = 162.5 mV .
AV = 36.3 dB
20 log AV = 36.3 AV = 65.
Vo = AV Vi = 65 2.5 mV = 162.5 mV .
dB in audio measurements
* When sound intensity is specified in dB, the reference pressure is Pref = 20 Pa

(our hearing threshold).
If the pressure corresponding to the sound being measured is P, we say that it is
20 log (P/Pref ) dB.

* Some interesting numbers:
mosquito 3 m away
0 dB

mosquito 3 m away
whisper
0 dB
20 dB

mosquito 3 m away
whisper
normal conversation
0 dB
20 dB
60 to 70 dB

mosquito 3 m away
whisper
normal conversation
noisy factory
0 dB
20 dB
60 to 70 dB
90 to 100 dB

mosquito 3 m away
whisper
normal conversation
noisy factory
loud thunder
0 dB
20 dB
60 to 70 dB
90 to 100 dB
110 dB

mosquito 3 m away
whisper
normal conversation
noisy factory
loud thunder
loudest sound human ear can tolerate
0 dB
20 dB
60 to 70 dB
90 to 100 dB
110 dB
120 dB

mosquito 3 m away
whisper
normal conversation
noisy factory
loud thunder
loudest sound human ear can tolerate
windows break
0 dB
20 dB
60 to 70 dB
90 to 100 dB
110 dB
120 dB
163 dB
Bode plots
Vi (s)
H(s)
Vo (s)
Bode plots
Vi (s)
H(s)
Vo (s)
* The transfer function of a circuit such as an amplifier or a filter is given by,

H(s) = Vo (s)/Vi (s), s = j.
e.g., H(s) =
K
K
=
1 + s
1 + j
Bode plots
Vi (s)
H(s)
Vo (s)

H(s) = Vo (s)/Vi (s), s = j.
e.g., H(s) =
K
K
=
1 + s
1 + j
* H(j) is a complex number, and a complete description of H(j) involves

(a) a plot of |H(j)| versus .
(b) a plot of H(j) versus .
Bode plots
Vi (s)
H(s)
Vo (s)

H(s) = Vo (s)/Vi (s), s = j.
e.g., H(s) =
K
K
=
1 + s
1 + j
* H(j) is a complex number, and a complete description of H(j) involves

(a) a plot of |H(j)| versus .
(b) a plot of H(j) versus .
* Bode gave simple rules which allow construction of the above Bode plots in an
approximate (asymptotic) manner.
A simple transfer function
R
(1/s C)
Vs ,
R + (1/s C)
1
1
=
,
H(s) =
1+s RC
1 + (j /0 )
1
.
0 =
RC
Vo =
Vs
Vo
R
(1/s C)
Vs ,
R + (1/s C)
1
1
=
,
H(s) =
1+s RC
1 + (j /0 )
1
.
0 =
RC
Vo =
Vs
Vo
* The circuit behaves like a low-pass filter.

For 0 , |H(j)| 1.
For 0 , |H(j)| 1/.
R
(1/s C)
Vs ,
R + (1/s C)
1
1
=
,
H(s) =
1+s RC
1 + (j /0 )
1
.
0 =
RC
Vo =
Vs
Vo

For 0 , |H(j)| 1.
For 0 , |H(j)| 1/.
* The magnitude and phase of H(j| are given by,

1
|H(j)| = p
, H(j) = tan1
.
0
1 + (/0 )2
R
(1/s C)
Vs ,
R + (1/s C)
1
1
=
,
H(s) =
1+s RC
1 + (j /0 )
1
.
0 =
RC
Vo =
Vs
Vo

For 0 , |H(j)| 1.
For 0 , |H(j)| 1/.

1
|H(j)| = p
, H(j) = tan1
.
0
1 + (/0 )2
* We are generally interested in a large variation in (several orders), and its
effect on |H| and H.
R
(1/s C)
Vs ,
R + (1/s C)
1
1
=
,
H(s) =
1+s RC
1 + (j /0 )
1
.
0 =
RC
Vo =
Vs
Vo

For 0 , |H(j)| 1.
For 0 , |H(j)| 1/.

1
|H(j)| = p
, H(j) = tan1
.
0
1 + (/0 )2
* We are generally interested in a large variation in (several orders), and its
effect on |H| and H.
* The magnitude (|H|) varies by orders of magnitude as well.
The phase (H) varies from 0 (for 0 ) to /2 (for 0 ).
A simple transfer function: magnitude
R
(1/s C)
Vs ,
R + (1/s C)
1
1
H(s) =
=
,
1+s RC
1 + (j /0 )
1
.
0 =
RC
Vo =
Vs
106
0
Frequency (rad/s)
Vo
100
100
10-1
10-1
10-2
10-2
10-3
106
0
Frequency (rad/s)
10-3
100
101
102
103
104
Frequency (rad/s)
105
106
R
(1/s C)
Vs ,
R + (1/s C)
1
1
H(s) =
=
,
1+s RC
1 + (j /0 )
1
.
0 =
RC
Vo =
Vs
106
0
Frequency (rad/s)
Vo
100
100
10-1
10-1
10-2
10-2
10-3
106
0
Frequency (rad/s)
10-3
100
101
102
103
104
105
Frequency (rad/s)
Since and |H(j )| vary by several orders of magnitude, a linear - or |H|-axis is not
appropriate log |H| is plotted against log .
106
R
(1/s C)
Vs ,
R + (1/s C)
1
1
H(s) =
=
,
1+s RC
1 + (j /0 )
1
0 =
.
RC
Vo =
C
Vs
Vo
0
0
10-1
20
|H|
|H| (dB)
100
10-2
10-3
100
40
101
102
103
104
Frequency (rad/s)
105
106
60
100
101
102
103
104
Frequency (rad/s)
105
106
R
(1/s C)
Vs ,
R + (1/s C)
1
1
H(s) =
=
,
1+s RC
1 + (j /0 )
1
0 =
.
RC
Vo =
C
Vs
Vo
0
0
10-1
20
|H|
|H| (dB)
100
10-2
10-3
100
40
101
102
103
104
105
106
60
100
101
Frequency (rad/s)
Note that the shape of the plot does not change.

|H| (dB) = 20 log |H| is simply a scaled version of log |H|.
102
103
104
Frequency (rad/s)
105
106
R
(1/s C)
Vs ,
R + (1/s C)
1
1
H(s) =
=
,
1+s RC
1 + (j /0 )
1
0 =
.
RC
Vo =
C
Vs
Vo
20
20
|H| (dB)
|H| (dB)
40
60
100
40
101
102
103
104
Frequency (rad/s)
105
106
60
100
101
102
103
104
Frequency (Hz)
105
106
R
(1/s C)
Vs ,
R + (1/s C)
1
1
H(s) =
=
,
1+s RC
1 + (j /0 )
1
0 =
.
RC
Vo =
C
Vs
Vo
20
20
|H| (dB)
|H| (dB)
40
60
100
40
101
102
103
104
105
106
60
100
Frequency (rad/s)
Since = 2 f , the shape of the plot does not change.
101
102
103
104
Frequency (Hz)
105
106
A simple transfer function: phase

R
(1/s C)
Vs ,
R + (1/s C)
1
1
H(s) =
=
,
1+s RC
1 + (j /0 )
1
0 =
.
RC
Vo =
Vs
Vo
0
0
90
106
0
Frequency (rad/s)
90
100
101
102
103
104
Frequency (rad/s)
105
106

R
(1/s C)
Vs ,
R + (1/s C)
1
1
H(s) =
=
,
1+s RC
1 + (j /0 )
1
0 =
.
RC
Vo =
Vs
Vo
0
0
90
106
0
Frequency (rad/s)
90
100
101
102
103
104
105
106
Frequency (rad/s)
* Since H = tan1 (/0 ) varies in a limited range (0 to 90 ), a linear axis

is appropriate for H.

R
(1/s C)
Vs ,
R + (1/s C)
1
1
H(s) =
=
,
1+s RC
1 + (j /0 )
1
0 =
.
RC
Vo =
Vs
Vo
0
0
90
106
0
Frequency (rad/s)
90
100
101
102
103
104
105
106
Frequency (rad/s)
* Since H = tan1 (/0 ) varies in a limited range (0 to 90 ), a linear axis

is appropriate for H.
* As in the magnitude plot, we use a log axis for , since we are interested in a
wide range of .
Construction of Bode plots
Consider H(s) =
K (1 + s/z1 )(1 + s/z2 ) (1 + s/zM )

.
(1 + s/p1 )(1 + s/p2 ) (1 + s/pN )
Consider H(s) =
K (1 + s/z1 )(1 + s/z2 ) (1 + s/zM )

.
(1 + s/p1 )(1 + s/p2 ) (1 + s/pN )
z1 , z2 , are called the zeros of H(s).

p1 , p2 , are called the poles of H(s).
(In addition, there could be terms like s, s 2 , in the numerator.)
We will assume, for simplicity, that the zeros (and poles) are real and distinct.
Construction of Bode plots involves
Consider H(s) =
K (1 + s/z1 )(1 + s/z2 ) (1 + s/zM )

.
(1 + s/p1 )(1 + s/p2 ) (1 + s/pN )

(a) computing approximate contribution of each pole/zero as a function .
Consider H(s) =
K (1 + s/z1 )(1 + s/z2 ) (1 + s/zM )

.
(1 + s/p1 )(1 + s/p2 ) (1 + s/pN )

(a) computing approximate contribution of each pole/zero as a function .
(b) combining the various contributions to obtain |H| and H versus .
Contribution of a pole: magnitude

p
20
asymptote 1
asymptote 2
|H| (dB)
exact
20
40
60
100
101
102
103
104
105
106
Frequency (rad/s)
Consider H(s) =
1
1
1
H(j) =
, |H(j)| = p
.
1 + s/p
1 + j (/p)
1 + (/p)2

p
20
asymptote 1
asymptote 2
|H| (dB)
exact
20
40
60
100
101
102
103
104
105
106
Frequency (rad/s)
Consider H(s) =
1
1
1
H(j) =
, |H(j)| = p
.
1 + s/p
1 + j (/p)
1 + (/p)2
Asymptote 1:
p: |H| 1, 20 log |H| = 0 dB.

p
20
asymptote 1
asymptote 2
|H| (dB)
exact
20
40
60
100
101
102
103
104
105
106
Frequency (rad/s)
Consider H(s) =
1
1
1
H(j) =
, |H(j)| = p
.
1 + s/p
1 + j (/p)
1 + (/p)2
Asymptote 1:
Asymptote 2:
p: |H| 1, 20 log |H| = 0 dB.

1
p
=
|H| = 20 log p 20 log (dB)
p: |H|
/p

p
20
asymptote 1
asymptote 2
|H| (dB)
exact
20
40
60
100
101
102
103
104
105
106
Frequency (rad/s)
Consider H(s) =
1
1
1
H(j) =
, |H(j)| = p
.
1 + s/p
1 + j (/p)
1 + (/p)2
Asymptote 1:
Asymptote 2:
p: |H| 1, 20 log |H| = 0 dB.

1
p
=
|H| = 20 log p 20 log (dB)
p: |H|
/p
Consider two values of : 1 and 10 1 .

|H|1 = 20 log p 20 log 1 (dB)
|H|2 = 20 log p 20 log (10 1 ) (dB)

p
20
asymptote 1
asymptote 2
|H| (dB)
exact
20
40
60
100
101
102
103
104
105
106
Frequency (rad/s)
Consider H(s) =
1
1
1
H(j) =
, |H(j)| = p
.
1 + s/p
1 + j (/p)
1 + (/p)2
Asymptote 1:
Asymptote 2:
p: |H| 1, 20 log |H| = 0 dB.

1
p
=
|H| = 20 log p 20 log (dB)
p: |H|
/p

|H|1 = 20 log p 20 log 1 (dB)
|H|2 = 20 log p 20 log (10 1 ) (dB)
1
|H|1 |H|2 = 20 log
= 20 dB.
10 1

p
20
asymptote 1
asymptote 2
|H| (dB)
exact
20
40
60
100
101
102
103
104
105
106
Frequency (rad/s)
Consider H(s) =
1
1
1
H(j) =
, |H(j)| = p
.
1 + s/p
1 + j (/p)
1 + (/p)2
Asymptote 1:
Asymptote 2:
p: |H| 1, 20 log |H| = 0 dB.

1
p
=
|H| = 20 log p 20 log (dB)
p: |H|
/p

|H|1 = 20 log p 20 log 1 (dB)
|H|2 = 20 log p 20 log (10 1 ) (dB)

1
|H|1 |H|2 = 20 log
= 20 dB.
10 1
|H| versus has a slope of 20 dB/decade.
Note that, at = p, the actual value of |H| is 1/ 2 (i.e., 3 dB).

Contribution of a pole: phase
p
asymptote 1
0
exact
asymptote 3
asymptote 2
90
100
101
102
103
104
105
106
Frequency (rad/s)
Consider H(s) =
1
1
1
=
H = tan
1 + s/p
1 + j (/p)
p
asymptote 1
0
exact
asymptote 3
asymptote 2
90
100
101
102
103
104
105
106
Frequency (rad/s)
Consider H(s) =
1
1
1
=
H = tan
1 + s/p
1 + j (/p)
Asymptote 1:
p (say, < p/10): H = 0.
p
asymptote 1
0
exact
asymptote 3
asymptote 2
90
100
101
102
103
104
105
106
Frequency (rad/s)
Consider H(s) =
1
1
1
=
H = tan
1 + s/p
1 + j (/p)
Asymptote 1:
p (say, < p/10): H = 0.
Asymptote 2:
p (say, > 10 p): H = /2.
p
asymptote 1
0
exact
asymptote 3
asymptote 2
90
100
101
102
103
104
105
106
Frequency (rad/s)
Consider H(s) =
1
1
1
=
H = tan
1 + s/p
1 + j (/p)
Asymptote 1:
p (say, < p/10): H = 0.
Asymptote 2:
p (say, > 10 p): H = /2.
Asymptote 3:
For p/10 < < 10 p , H is assumed to vary linearly with log

at = p, H = /4 (which is also the actual value of H).
Contribution of a zero: magnitude

z
60
|H| (dB)
40
20
exact
0
asymptote 2
asymptote 1
20
100
101
102
103
104
105
106
Frequency (rad/s)
Consider H(s) = 1 + s/z H(j) = 1 + j (/z) , |H(j)| =
1 + (/z)2 .

z
60
|H| (dB)
40
20
exact
0
asymptote 2
asymptote 1
20
100
101
102
103
104
105
106
Frequency (rad/s)

Asymptote 1:
p: |H| 1, 20 log |H| = 0 dB.
1 + (/z)2 .

z
60
|H| (dB)
40
20
exact
0
asymptote 2
asymptote 1
20
100
101
102
103
104
105
106
Frequency (rad/s)

Asymptote 1:
Asymptote 2:
1 + (/z)2 .
p: |H| 1, 20 log |H| = 0 dB.
|H| = 20 log 20 log z (dB)

z
p: |H|

z
60
|H| (dB)
40
20
exact
0
asymptote 2
asymptote 1
20
100
101
102
103
104
105
106
Frequency (rad/s)

Asymptote 1:
Asymptote 2:
1 + (/z)2 .
p: |H| 1, 20 log |H| = 0 dB.
|H| = 20 log 20 log z (dB)

z
p: |H|
|H|1 = 20 log 1 20 log z (dB)

|H|2 = 20 log (10 1 ) 20 log z (dB)

z
60
|H| (dB)
40
20
exact
0
asymptote 2
asymptote 1
20
100
101
102
103
104
105
106
Frequency (rad/s)

Asymptote 1:
Asymptote 2:
1 + (/z)2 .
p: |H| 1, 20 log |H| = 0 dB.
|H| = 20 log 20 log z (dB)

z
p: |H|
|H|1 = 20 log 1 20 log z (dB)

|H|2 = 20 log (10 1 ) 20 log z (dB)
1
|H|1 |H|2 = 20 log
= 20 dB.
10 1

z
60
|H| (dB)
40
20
exact
0
asymptote 2
asymptote 1
20
100
101
102
103
104
105
106
Frequency (rad/s)

Asymptote 1:
Asymptote 2:
1 + (/z)2 .
p: |H| 1, 20 log |H| = 0 dB.
|H| = 20 log 20 log z (dB)

z
p: |H|
|H|1 = 20 log 1 20 log z (dB)

|H|2 = 20 log (10 1 ) 20 log z (dB)
1
|H|1 |H|2 = 20 log
= 20 dB.
10 1
|H| versus has a slope of +20 dB/decade.
Note that, at = z, the actual value of |H| is 2 (i.e., 3 dB).

Contribution of a zero: phase
90
exact
asymptote 2
asymptote 3
asymptote 1
0
100
101
102
103
104
105
106
Frequency (rad/s)
Consider H(s) = 1 + s/z = 1 + j (/z) H = tan1
90
exact
asymptote 2
asymptote 3
asymptote 1
0
100
101
102
103
104
105
106
Frequency (rad/s)

Asymptote 1:
z (say, < z/10): H = 0.
90
exact
asymptote 2
asymptote 3
asymptote 1
0
100
101
102
103
104
105
106
Frequency (rad/s)

Asymptote 1:
z (say, < z/10): H = 0.
Asymptote 2:
z (say, > 10 z): H = /2.
90
exact
asymptote 2
asymptote 3
asymptote 1
0
100
101
102
103
104
105
106
Frequency (rad/s)
Asymptote 1:
z (say, < z/10): H = 0.
Asymptote 2:
z (say, > 10 z): H = /2.
Asymptote 3:
For z/10 < < 10 z , H is assumed to vary linearly with log

at = z, H = /4 (which is also the actual value of H).
Contribution of K (constant), s, and s 2

For H(s) = K , 20 log |H| = 20 log K (a constant), and H = 0 .

120
|s2 | (dB)
|s| (dB)
60
40
40
20
100
80
101
102
Frequency (rad/s)
103
100
101
102
Frequency (rad/s)
103

120
|s2 | (dB)
|s| (dB)
60
40
40
20
100
80
101
102
103
Frequency (rad/s)
For H(s) = s, i.e., H(j) = j, |H| = .
100
101
102
Frequency (rad/s)
103

120
|s2 | (dB)
|s| (dB)
60
40
40
20
100
80
101
102
103
Frequency (rad/s)
For H(s) = s, i.e., H(j) = j, |H| = .

20 log |H| = 20 log ,
100
101
102
Frequency (rad/s)
103

120
|s2 | (dB)
|s| (dB)
60
40
40
20
100
80
101
102
Frequency (rad/s)
103
100
101
102
103
Frequency (rad/s)
For H(s) = s, i.e., H(j) = j, |H| = .

20 log |H| = 20 log ,
i.e., a straight line in the |H| (dB)-log plane with a slope of 20 dB/decade,
passing through (1, 0).

120
|s2 | (dB)
|s| (dB)
60
40
40
20
100
80
101
102
Frequency (rad/s)
103
100
101
102
103
Frequency (rad/s)
For H(s) = s, i.e., H(j) = j, |H| = .

20 log |H| = 20 log ,
H = /2 (irrespective of ).

120
|s2 | (dB)
|s| (dB)
60
40
40
20
100
80
101
102
103
100
Frequency (rad/s)
101
102
103
Frequency (rad/s)
For H(s) = s, i.e., H(j) = j, |H| = .

20 log |H| = 20 log ,
For H(s) = s 2 , i.e., H(j) = 2 , |H| = 2 .

120
|s2 | (dB)
|s| (dB)
60
40
40
20
100
80
101
102
103
100
Frequency (rad/s)
101
102
103
Frequency (rad/s)
For H(s) = s, i.e., H(j) = j, |H| = .

20 log |H| = 20 log ,
For H(s) = s 2 , i.e., H(j) = 2 , |H| = 2 .
20 log |H| = 40 log ,

120
|s2 | (dB)
|s| (dB)
60
40
40
20
100
80
101
102
Frequency (rad/s)
103
100
101
102
103
Frequency (rad/s)
For H(s) = s, i.e., H(j) = j, |H| = .

20 log |H| = 20 log ,
For H(s) = s 2 , i.e., H(j) = 2 , |H| = 2 .
20 log |H| = 40 log ,

120
|s2 | (dB)
|s| (dB)
60
40
40
20
100
80
101
102
Frequency (rad/s)
103
100
101
102
103
Frequency (rad/s)
For H(s) = s, i.e., H(j) = j, |H| = .

20 log |H| = 20 log ,
For H(s) = s 2 , i.e., H(j) = 2 , |H| = 2 .
20 log |H| = 40 log ,
H = (irrespective of ).
Combining different terms
Consider H(s) = H1 (s) H2 (s).

Magnitude:
|H(j)| = |H1 (j)| |H2 (j)|.
20 log |H| = 20 log |H1 | + 20 log |H2 |.

Magnitude:
|H(j)| = |H1 (j)| |H2 (j)|.
20 log |H| = 20 log |H1 | + 20 log |H2 |.
In the Bode magnitude plot, the contributions due to H1 and H2 simply get added.

Magnitude:
|H(j)| = |H1 (j)| |H2 (j)|.
20 log |H| = 20 log |H1 | + 20 log |H2 |.
Phase:
H1 (j) and H2 (j) are complex numbers.
At a given , let H1 = K1 = K1 e j , and H2 = K2 = K2 e j .
Then, H1 H2 = K1 K2 e j(+) = K1 K2 ( + ) .
i.e., H = H1 + H2 .

Magnitude:
|H(j)| = |H1 (j)| |H2 (j)|.
20 log |H| = 20 log |H1 | + 20 log |H2 |.
Phase:
Then, H1 H2 = K1 K2 e j(+) = K1 K2 ( + ) .
i.e., H = H1 + H2 .
In the Bode phase plot, the contributions due to H1 and H2 also get added.

Magnitude:
|H(j)| = |H1 (j)| |H2 (j)|.
20 log |H| = 20 log |H1 | + 20 log |H2 |.
Phase:
Then, H1 H2 = K1 K2 e j(+) = K1 K2 ( + ) .
i.e., H = H1 + H2 .
In the Bode phase plot, the contributions due to H1 and H2 also get added.
The same reasoning applies to more than two terms as well.
Combining different terms: example
Consider H(s) =
10 s
.
(1 + s/102 ) (1 + s/105 )
Consider H(s) =
10 s
.
(1 + s/102 ) (1 + s/105 )
Let H(s) = H1 (s) H2 (s) H3 (s) H4 (s) , where

H1 (s) = 10 ,
H2 (s) = s ,
1
, p1 = 102 rad/s,
1 + s/p1
1
H4 (s) =
, p2 = 105 rad/s.
1 + s/p2
H3 (s) =
Consider H(s) =
10 s
.
(1 + s/102 ) (1 + s/105 )
Let H(s) = H1 (s) H2 (s) H3 (s) H4 (s) , where

H1 (s) = 10 ,
H2 (s) = s ,
1
, p1 = 102 rad/s,
1 + s/p1
1
H4 (s) =
, p2 = 105 rad/s.
1 + s/p2
H3 (s) =
We can now plot the magnitude and phase of H1 , H2 , H3 , H4 individually versus

and then simply add them to obtain |H| and H.
Magnitude plot (|H| in dB)
80
H1 (s) = 10
60
40
20
0
100
101
102
103
104
105
Frequency (rad/s)
106
107
80
80
H1 (s) = 10
H2 (s) = s
60
60
40
40
20
20
100
101
102
103
104
105
Frequency (rad/s)
106
107
100
101
102
103
104
105
Frequency (rad/s)
106
107
80
80
H1 (s) = 10
H2 (s) = s
60
60
40
40
20
20
100
101
102
103
104
105
106
107
Frequency (rad/s)
0
20
60
100
H3 (s) =
101
1
1 + s/102
102
103
104
105
Frequency (rad/s)
100
101
102
103
104
105
Frequency (rad/s)
20
40
106
107
106
107
80
80
H1 (s) = 10
H2 (s) = s
60
60
40
40
20
20
100
101
102
103
104
105
106
107
100
101
Frequency (rad/s)
20
20
20
60
100
H3 (s) =
101
40
1
1 + s/102
102
103
104
105
Frequency (rad/s)
103
104
105
106
107
106
107
Frequency (rad/s)
20
40
102
106
107
60
100
H4 (s) =
101
1
1 + s/105
102
103
104
105
Frequency (rad/s)
80
80
H1 (s) = 10
80
H2 (s) = s
60
60
60
40
40
40
20
20
20
100
101
102
103
104
105
106
107
100
101
Frequency (rad/s)
20
20
20
60
100
H3 (s) =
101
40
1
1 + s/102
102
103
104
105
Frequency (rad/s)
103
104
105
106
107
Frequency (rad/s)
20
40
102
106
107
60
100
H4 (s) =
101
103
104
105
Frequency (rad/s)
100
101
10 s
(1 + s/102 ) (1 + s/105 )
102
103
104
105
Frequency (rad/s)
1
1 + s/105
102
H(s) =
106
107
106
107
80
80
H1 (s) = 10
80
H2 (s) = s
60
60
60
40
40
40
20
20
20
100
101
102
103
104
105
106
107
100
101
Frequency (rad/s)
102
103
104
105
106
107
H(s) =
100
101
10 s
(1 + s/102 ) (1 + s/105 )
102
Frequency (rad/s)
103
104
105
106
107
Frequency (rad/s)
20
20
80
60
20
20
40
approx.
exact
40
60
H3 (s) =
100
101
40
1
1 + s/102
102
103
104
105
Frequency (rad/s)
106
107
60
100
H4 (s) =
101
20
1
1 + s/105
102
103
104
105
Frequency (rad/s)
106
107
100
H(s) =
101
10 s
(1 + s/102 ) (1 + s/105 )
102
103
104
105
Frequency (rad/s)
106
107
Phase plot
90
45
0
45
H1 (s) = 10
90
100
101
102
103
104
105
Frequency (rad/s)
106
107
Phase plot
90
90
45
45
45
45
H1 (s) = 10
90
100
H2 (s) = s
90
101
102
103
104
105
Frequency (rad/s)
106
107
100
101
102
103
104
105
Frequency (rad/s)
106
107
Phase plot
90
90
45
45
45
45
H1 (s) = 10
90
100
101
102
103
104
105
106
107
Frequency (rad/s)
H3 (s) =
45
1
1 + s/102
0
45
90
101
102
103
104
105
Frequency (rad/s)
100
101
102
103
104
105
Frequency (rad/s)
90
100
H2 (s) = s
90
106
107
106
107
Phase plot
90
90
45
45
45
45
H1 (s) = 10
90
100
H2 (s) = s
90
101
102
103
104
105
106
107
100
101
Frequency (rad/s)
45
90
1 + s/102
45
0
45
90
100
104
105
H4 (s) =
0
45
103
106
107
Frequency (rad/s)
90
H3 (s) =
102
1
1 + s/105
90
101
102
103
104
105
Frequency (rad/s)
106
107
100
101
102
103
104
105
Frequency (rad/s)
106
107
Phase plot
90
90
90
45
45
45
45
45
45
H1 (s) = 10
90
100
102
103
104
105
106
107
100
90
101
Frequency (rad/s)
H3 (s) =
45
90
1 + s/102
45
0
45
90
100
103
104
105
H4 (s) =
0
45
102
106
107
Frequency (rad/s)
90
102
103
104
105
Frequency (rad/s)
106
107
100
101
102
103
104
1
1 + s/105
105
Frequency (rad/s)
100
101
102
103
104
105
Frequency (rad/s)
90
101
10 s
(1 + s/102 ) (1 + s/105 )
H2 (s) = s
90
101
H(s) =
106
107
106
107
Phase plot
90
90
90
45
45
45
45
45
45
H1 (s) = 10
90
101
102
103
104
105
106
107
100
90
101
Frequency (rad/s)
H3 (s) =
45
102
103
104
105
106
100
107
90
H4 (s) =
1 + s/102
45
90
1
1 + s/105
45
45
90
100
101
102
103
104
105
Frequency (rad/s)
106
107
100
103
H(s) =
45
90
102
104
105
106
107
Frequency (rad/s)
0
45
101
Frequency (rad/s)
90
10 s
(1 + s/102 ) (1 + s/105 )
H2 (s) = s
90
100
H(s) =
10 s
(1 + s/102 ) (1 + s/105 )
exact
approx.
90
101
102
103
104
105
Frequency (rad/s)
106
107
100
101
102
103
104
105
Frequency (rad/s)
106
107
How good are the approximations?
* As we have seen, the contribution of a pole to the magnitude and phase plots is
well represented by the asymptotes when p or p (similarly for a zero).
* Near = p (or = z), there is some error.
* If two poles p1 and p2 are close to each other (say, separated by less than a
decade in ), the error becomes larger (next slide).
* If two poles p1 and p2 are close to each other (say, separated by less than a
decade in ), the error becomes larger (next slide).
* When the poles and zeros are not sufficiently separated, the Bode approximation
should be used only for a rough estimate, follwed by a numerical calculation.
However, even in such cases, it does give a good idea of the asymptotic
magnitude and phase plots, which is valuable in amplifier design.
Consider H(s) =
10 s
.
(1 + s/p1 ) (1 + s/p2 )
p1 = 102
p2 = 104
approx.
|H| (dB)
60
exact
40
20
100
101
102
103
104
Frequency (rad/s)
105
106
Consider H(s) =
10 s
.
(1 + s/p1 ) (1 + s/p2 )
p1 = 102
p2 = 104
approx.
|H| (dB)
60
p1 = 102
p2 = 103
exact
40
20
100
101
102
103
104
Frequency (rad/s)
105
106 100
101
102
103
104
Frequency (rad/s)
105
106
Consider H(s) =
10 s
.
(1 + s/p1 ) (1 + s/p2 )
p1 = 102
p2 = 104
approx.
|H| (dB)
60
p1 = 102
p2 = 103
p1 = 102
p2 = 5 102
exact
40
20
100
101
102
103
104
Frequency (rad/s)
105
106 100
101
102
103
104
Frequency (rad/s)
105
106 100
101
102
103
104
Frequency (rad/s)
105
106
Consider H(s) =
10 s
.
(1 + s/p1 ) (1 + s/p2 )
p1 = 102
p2 = 104
90
45
0
45
90
100
101
102
103
104
Frequency (rad/s)
105
106
Consider H(s) =
10 s
.
(1 + s/p1 ) (1 + s/p2 )
p1 = 102
p2 = 103
p1 = 102
p2 = 104
90
45
0
45
90
100
101
102
103
104
Frequency (rad/s)
105
106 100
101
102
103
104
Frequency (rad/s)
105
106
Consider H(s) =
10 s
.
(1 + s/p1 ) (1 + s/p2 )
p1 = 102
p2 = 103
p1 = 102
p2 = 104
90
45
p1 = 102
p2 = 5 102
0
45
90
100
101
102
103
104
Frequency (rad/s)
105
106 100
101
102
103
104
Frequency (rad/s)
105
106 100
101
102
103
104
Frequency (rad/s)
105
106

Ee101 Bode 1 PDF

Hochgeladen von

Dokumentinformationen

Originalbeschreibung:

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Ee101 Bode 1 PDF

Hochgeladen von

Copyright:

Verfügbare Formate

EE101: Bode plots

M. B. Patil, IIT Bombay

What is deciBel (dB)?

* The unit dB is used to represent quantities on a logarithmic scale.

M. B. Patil, IIT Bombay

What is deciBel (dB)?

* The unit dB is used to represent quantities on a logarithmic scale.

M. B. Patil, IIT Bombay

What is deciBel (dB)?

* The unit dB is used to represent quantities on a logarithmic scale.

M. B. Patil, IIT Bombay

What is deciBel (dB)?

* The unit dB is used to represent quantities on a logarithmic scale.

M. B. Patil, IIT Bombay

What is deciBel (dB)?

* The unit dB is used to represent quantities on a logarithmic scale.

M. B. Patil, IIT Bombay

What is deciBel (dB)?

* The unit dB is used to represent quantities on a logarithmic scale.

M. B. Patil, IIT Bombay

What is deciBel (dB)?

* The unit dB is used to represent quantities on a logarithmic scale.

M. B. Patil, IIT Bombay

What is deciBel (dB)?

* The unit dB is used to represent quantities on a logarithmic scale.

M. B. Patil, IIT Bombay

What is deciBel (dB)?

* dB is a unit that describes a quantity, on a log scale, with respect to a reference

M. B. Patil, IIT Bombay

What is deciBel (dB)?

* dB is a unit that describes a quantity, on a log scale, with respect to a reference

M. B. Patil, IIT Bombay

What is deciBel (dB)?

* dB is a unit that describes a quantity, on a log scale, with respect to a reference

M. B. Patil, IIT Bombay

What is deciBel (dB)?

* dB is a unit that describes a quantity, on a log scale, with respect to a reference

V1 = 10 log (1.2 V /1 mV )2 = 20 log 1.2/103 = 61.6 dBm.

M. B. Patil, IIT Bombay

What is deciBel (dB)?

* dB is a unit that describes a quantity, on a log scale, with respect to a reference

V1 = 10 log (1.2 V /1 mV )2 = 20 log 1.2/103 = 61.6 dBm.

M. B. Patil, IIT Bombay

Given Vi = 2.5 mV and AV = 36.3 dB,

M. B. Patil, IIT Bombay

Given Vi = 2.5 mV and AV = 36.3 dB,

M. B. Patil, IIT Bombay

Given Vi = 2.5 mV and AV = 36.3 dB,

Vo = 7.96 + 36.3 = 44.22 dBm

M. B. Patil, IIT Bombay

Given Vi = 2.5 mV and AV = 36.3 dB,

Vo = 7.96 + 36.3 = 44.22 dBm

M. B. Patil, IIT Bombay

Given Vi = 2.5 mV and AV = 36.3 dB,

Vo = 7.96 + 36.3 = 44.22 dBm

M. B. Patil, IIT Bombay

Given Vi = 2.5 mV and AV = 36.3 dB,

Vo = 7.96 + 36.3 = 44.22 dBm

M. B. Patil, IIT Bombay

Given Vi = 2.5 mV and AV = 36.3 dB,

Vo = 7.96 + 36.3 = 44.22 dBm

M. B. Patil, IIT Bombay