Bode Plot

Why Bode Plot?

• The determination of frequency response by traditional methods (like the analytical
methods) is tedious, when transfer function has several poles and zeros.

• The difficulty involved in the computation in frequency domain can be considerably

reduced by utilizing graphical technique in logarithmic co-ordinate system for the
frequency response analysis for the system.

• Thus, a logarithmic plot of transfer function in frequency domain is called BODE PLOT.

• Bode Plot consists of two plots/graphs

1) G ( j ) in db Vs log 
2) G ( j ) in deg Vs log 
 LOGARITHMIC UNITS
• The magnitude of the sinusoidal transfer function (or in other words transfer function in
frequency domain) can be expressed in db (decibel), which is a logarithmic unit as
20log10 G ( j ) db
• This magnitude is called log magnitude.

Advantage:
• The products of the factors in G ( j ) becomes additive terms because of the use of log.

UNITS TO EXPRESS FREQUENCY BANDS / RATIOS IN LOG

SCALE.

 OCTAVE

 DECADE
OCTAVE:
An octave is a frequency band from (say) f1 to f2 , where
(f2 / f1) = 2
Thus the frequency band from 1 to 2 c/s is one octave, or band from 2 to 4 c/s is also one octave.
DECADE:
If (f2 / f1) = 10, it is called DECADE

SIMPLE CALCULATIONS AND CONVERSIONS

• No. of octaves in a frequency range from f1 to f2 = log (f2 / f1) / log 2
= 3.32 log (f2 / f1) oct.

• No. of decades in a frequency range from f1 to f2 = log (f2 / f1) / log10 10

= log (f2 / f1) decade
• No. of octaves = 1 / log 2 = 1/ 0.301 decade

• 20* N db/decade = (20*N) (0.301) db/ octave = 6*N db/octave

 BODE PLOT PRELIMINARIES
For bode plot the sinusoidal transfer function G ( j ) must be expressed in time-constant form,
as shown in the expression below,

Let us consider a generalized form of G ( j )

2 1
K ( j ) n (1  jT1 )[1  ( j )  ( j ) 2
].........
n n 2
G ( j ) 
2 1
( j ) n (1  jTa )[1  ( j )  2 ( j ) 2 ].......
n n
2
   
K ( j ) n (1  jT1 )[1   2   j 2  ]........
 n   n 
G ( j )  2
   
( j ) n (1  jTa )[1   2   j 2  ].......
 n   n 
We find from above that the numerator and denominator have 4 basic possible types of
factors :
 1. Constant gain K

2. Poles or zeros at origin ( j ) n

3. Simple poles or zeros (1  jT )1

For repeated poles or zeros (1  jT )

r

2
     1
4. Quadratic poles or zeros [1   2   j 2  ]
 n   n 
2
     r
For repetitions [1   2   j 2  ]
 n   n 

Note: The transfer function of the system may have all the factors in it or combination of
some of the factors.
 OBJECTIVE:
• We are to draw the bode plot by straight line asymptotes or in other words we will learn to
draw approximate function plot.

• The exact plot can be obtained from the approximate plot via a technique called error
corrections at each corner frequencies.

STEP- BY - STEP PROCEDURE TO DRAW BODE PLOT

 First we learn to draw the bode plot for each individual factors as discussed above.

 Secondly we will learn to draw bode plot for a sinusoidal transfer function which
consists of combination of some of the factors of G ( j ) discussed above.
 BODE PLOT FOR INDIVIDUAL FACTORS OF G ( j )
1. CONSTANT GAIN FACTOR (K)
G ( j )  K
The gain “K” is frequency independent term. Finding the log magnitude,
G ( j )  20 (log10 K ) db  Constant

G ( j )  0 degrees
G ( j )
G ( j )
+180

+90
20 (log10 K ) db  (log scale)
0
-90
-180

0.1 1.0 10
2. POLES OR ZEROS AT ORIGIN
G ( j )  ( j )  n
where n = no. of repetations
+ n indicates no. of zeros at origin
- n indicates no. of poles at origin
1
G ( j )  ( j ) 1  here n  1  pole
j
G( j)  ( j)1  j here n  1 zero
Finding log magnitude and angle of G ( j )
1
If G ( j )  ( j )  n 
G ( j )  20 log   1 db ( j ) n
  20 log   db G ( j )  20 log    n db
  20  n log   db
Per unit change in log 
 Unit change in log  means:
 2 
log   1
 1 
 log 2  log 1  log10 10
2
  10
1
 2  10 1
So, using the above concept, we can write
G ( j )   20 db / decade  for single pole at origin
and
G ( j )   20  n db / decade  for multiple poles at origin
Interpretation : This means, when G ( j ) db is plotted against  it is a straight line with
negative slope of 20*n db / decade for poles and positive slope of 20*n db / decade for
zeros.
 Angle plot for G ( j )   j 
n

G ( j )   n  90 ; + ve for zeros and - ve for poles

If n 1 then If n  2 then
G ( j )   90 for zeros G ( j )   180 for zeros
G ( j )   90 for poles G ( j )   180 for poles

+ 40

+ 20
G ( j ) +20 db/ decade for single zero
0 at origin

- 20

- 40
- 20 db/ decade for single pole
at origin

0.1 1 10
 The angle plot is constant +/- 90 degrees for a single zero/ pole at origin
as the frequency  varies.

G ( j )
For zeros
180

 90

0.1 1 10 
90

180

For poles
3. SIMPLE POLES OR ZEROS
1
G ( j )   (1  jT )  n  This is a repeated simple pole
(1  jT ) n

If n = 1 , then
1
G ( j )   (1  jT ) 1  This is a non-repeated simple pole
(1  jT )
Similarly,
G ( j )  (1  jT ) n   This is a repeated simple zero
If n = 1 , then
G ( j )  (1  jT )1  This is a non-repeated simple zero

We consider the case of non-repeated simple pole first

1
G ( j )   (1  jT ) 1
(1  jT )
 1  Unlike other factors discussed so far it has
G ( j ) in db  20 log10  
 1 2 T 2



two frequency regions in it
1 Low frequency region
 

 20 log10 1   T
2 2 2
If  T  1
  10 log10 (1   T )
2 2

High frequency region

If  T  1
In low frequency region In high frequency region
G( j )  10log10 (1   2 T 2 ) G ( j )  10log10 (1   2 T 2 )
 10log10 1  10log10 ( 2 T 2 )
 0 db   20log10 ( T ) db

The G ( j ) db plot is a st. line The G ( j ) db plot is a st. line of

of slope 0 db/decade for 0   T 1 slope -20 db/decade, passing
through 0db at  T 1 for 1  T 

1
Important Note: The frequency   or  T  1
at which two
T
asymptotes (high and low frequency) meet is called CORNER
FREQUENCY. It divides the plot into 2 regions (high and low).
 High frequency asymptote for simple zero-
+ 20 db/decade st line passing through corner
G ( j ) db
frequency  T 1
+ 40
Low frequency asymptote

+ 20

- 20

- 40


0.1 1 10
High frequency asymptote for simple pole -
1 - 20 db/decade st line passing through corner
Corner frequency   
T frequency  T 1
 Angle for the simple pole (non-repeated) Angle for thesimple pole (repeated)
1
G ( j )  is given by G ( j ) 
1
1  j T  (1  j T ) n
is given by

 T   T 
 G ( j )    tan  1    G ( j )    n  tan  1 
 1  
 1 
G ( j )
Phase plot for simple zero
180

 90
 45
0.1 1 10 
45
90

180 Phase plot for simple pole

4. COMPLEX CONJUGATE PAIR OF POLES OR ZEROS
1
   2     
G ( j )  1     j 2    , Let    u  normalized frequency
  n   n    n 
1
 G ( j ) 
(1  u 2 ) 2  (2  u ) 2
1

 G ( j ) db  20 log [ (1  u )  (2  u ) ]
2 2 2 2 Corner frequency for this T.F is u  1
 10 log[ (1  u 2 ) 2  (2  u ) 2 ] 
  1    n
In Low frequency region u  1
n
G ( j )   10 log1
 0 db
In High frequency region u  1
G ( j )   10 log (u 4  4  2 u 2 );as   1we know and u 2 < u 4
so second term is neglected
thus,
= -10 log u 4
G ( j ) db   40 log u
 Log – Magnitude Plot
G ( j ) db

+ 40
Low frequency asymptotes + 40 db/decade (for zeros)
+ 20
High frequency asymptotes
0

- 40 db/decade (for poles)

--20
40

- 60 u
0.1 1 10
 Phase Calculation
 2 u 
 G ( j )   tan 1    for poles For non-repeated complex
 1 u 
2

 2 u  conjugate pairs of poles or zeros

 G ( j )   tan 1    for zeros
 1 u 
2


The phase plot can be obtained by varying u
n
 SUMMARY
In short the procedure for drawing “BODE PLOT” for a given system with combination of various
factors of G( j ) discussed above can summarized as:

1. Write the sinusoidal transfer function in time-constant form.

2. Identify the corner frequency (C.F) associated with each factor, if it exists.
3. Arrange all the factors in terms of ascending corner frequencies.
3. Knowing the C.F, draw the log magnitude plot. This plot consists of st. line segments with line
slope changing at each C.F.
+20*n db/decade for simple zero and -20*n db/decade for simple pole.
+40*n db/decade -40*n db/decade for complex conjugate pair of zeros and poles
respectively.
***Add slopes of all the factors algebraically at each C.F to get resultant magnitude plot
for the given T.F.
4. Draw phase angle curves for each factor and add them algebraically to get phase plot for the
given T.F.
 PROBLEM 1
Sketch the Bode plot for a unity feedback system characterized by the loop transfer
function
K (1  0.2 s) (1  0.025s)
G( s)  3
s (1  0.001 s) (1  0.005 s)

Show the system is conditionally stable. Find the range of ‘K’ for which the system is
stable.
SOLUTION:
Assume K 1, and alsosubstitute s  j in the given G(s)
(1  0.2 j ) (1  0.025 j )
G ( j ) 
( j )3 (1  0.001 j ) (1  0.005 j )

Log-magnitude characteristics for various factors

Factors C.F (rad/sec) Asymptotic log-mag. Characteristics

1 None -20 X 3 db/decade = -60db/decade. A st. line of

constant slope – 60 db/decade passing through
( j )3
 =1
1
(1  0.2 j )   5 A st. line of slope 0 db/decade upto C.F. Than a
0.2 st. line of slope + 20 db/decade after C.F.

(1  0.025 j )   40
A st. line of slope 0 db/decade upto C.F. Than a
st. line of slope + 20 db/decade after C.F.
1
(1  0.005 j )
  200 A st. line of slope 0 db/decade upto C.F. Than a
st. line of slope - 20 db/decade after C.F.
1
(1  0.001 j )
  1000 A st. line of slope 0 db/decade upto C.F. Than a
st. line of slope - 20 db/decade after C.F.
Phase Plot Calculation So, resultant   is given by:
  = - 270 + tan  1  0.2    tan  1  0.025  
 tan  1  0.005    tan  1  0.001 
 1  270
 Phase angle
(rad/sec) degrees
 2 tan  1  0.2  
0.1 - 268
0.8 -260
 3 tan 1
 0.025   1 -257
3 -236

 4  tan  1  0.005  
10 -198
15 -183
30 -163
 5  tan 1
 0.001  60 -148
100 -147
300 -172
400 -182
 + 60
+ 40 - 60 db/decade
+ 20
0
- 40 db/decade
- 40
-60 - 20 db/decade
- 80 - 40 db/decade
- 100

1 10 100 1000
- 60 db/decade
 G ( j )

90

180

 270


0.1 1 10
For Phase Plot

Note: Draw both the log-magnitude and phase angle plot in the same graph paper with
common frequency axis.
 RELATIVE STABILITY ANALYSIS FROM BODE
PLOT
For carrying out the relative stability analysis for a given system using
Bode Plot, two important terms will be introduced first

GAIN MARGIN

PHASE MARGIN

They are required for knowing the margin or limit of stability for a
system. It gives the relative stability measure of the system..
GAIN MARGIN: The gain margin is a factor by which the gain of a
stable system is allowed to increase before the system reaches
instability.

PHASE MARGIN: The phase margin of a stable system is the amount

of additional phase lag required to bring the system to the point of
instability.

HOW GAIN AND PHASE MARGINS ARE CALCULATED

FROM BODE PLOT ? ?
Note: For these calculations, it is customary to plot log-magnitude
and phase plots in the same semi-log graph paper.
 For Bode plot GAIN MARGIN is calculated using phase plot. We find phase cross over
frequency in the phase plot, which is defined as:
the frequency at which phase plot crosses “– 180” degree line, denoted by  .

Now corresponding to phase cross over frequency in the phase plot, we look for the amount
of gain in db in the log-magnitude plot.

GAIN MARGIN will be amount of gain required to raise the the magnitude curve to 0 db.

So, if G.M. is positive, system is stable.

If G.M. is negative than system is unstable.
If G.M. is 0 (zero) than system is marginally stable.
 For BODE PLOT phase margin is calculated from the log-magnitude plot. We find
gain cross over frequency, which is defined as
the frequency at which the log-magnitude plot crosses 0 db line, denoted by  g .

Now corresponding to gain cross over frequency in the gain plot, we look for the amount
of phase angle in degrees in the phase plot to find the Phase margin (P.M).

P.M .    ( 180 )
 180  
If P.M. comes out to be positive than the system is stable, else unstable.
 ALTERNATIVE TECHNIQUE

Analytical method to calculate the following parameters

GAIN CROSS OVER FREQUENCY
PHASE CROSS OVER FREQUENCY
GAIN MARGIN
PHASE MARGIN
Gain Margin (G.M) Calculation
G ( j ) H ( j ) A
Finding  (Phase cross over frequency) at   

1
Im G ( j ) H ( j )  0 G.M  20 log   db
 A

Finding  g (Gain cross over frequency) Phase Margin (P.M) Calculation

G ( j ) H ( j )  1 P.M  180   G ( j ) H ( j )  
g
G ( j ) g
0 db  Here G.M is positive
G.M
Here P.M is positive

G ( j )
g  
P.M
180   System is stable
G ( j )
g
G.M
0 db 
Here G.M is negative

Here P.M is negative

G ( j )
 g  
180 
P.M System is unstable
G ( j ) g
0 db 
Here G.M is zero

Here P.M is zero

G ( j )
g  
180 
System is marginally stable

 2n

s( s  2n )

c ( j ) n2 r  n 1  2 2

R( j ) ( j ) 2  j 2n  n2
1
 Mr 
put u   normalized frequency 2 1   2
n
1  2 2
M (j ) =
1 r   tan 1

(1-u2 )  j 2 u 
1
M (j )  1

(1-u2 )  (2 u ) 2 B.W  n 1  2   2  4  4

2 2

4 2

2 u
M (j ) = - tan -1
1  u2