10/28/2015

lesson15et438a.pptx 1

LESSON 15: BODE PLOTS

OF TRANSFER FUNCTIONS
ET 438a Automatic Control Systems Technology

Learning Objectives
2

After this presentation you will be able to:

 Compute the magnitude of a transfer function for a given

radian frequency.
 Compute the phase shift of a transfer function for a given
radian frequency.
 Construct a Bode plot that shows both magnitude and
phase shift as functions of transfer function input frequency
 Use MatLAB instructions it produce Bode plots of transfer
functions.

lesson15et438a.pptx

1
 10/28/2015

Construction of Bode Plots

3

Bode plots consist of two individual graphs: a) a semilog plot of gain vs frequency
b) a semilog plot of phase shift vs frequency. Frequency is the logarithmic axis on
both plots.

Bode plots of transfer functions give the frequency response of a control system

To compute the points for a Bode Plot:

1) Replace Laplace variable, s, in transfer function with jw

2) Select frequencies of interest in rad/sec (w=2pf)
3) Compute magnitude and phase angle of the resulting
complex expression.

lesson15et438a.pptx

Construction of Bode Plots

4

Bode plot calculations- magnitude/phase

X(jw) Y(jw)
Transfer
Function

Y( jw)
G ( jw)  Where for a given frequency
Gain: X( jw)
Y  yo Phase:   o  i
dB  20  log( G ( jw) )
X  xi

To find the magnitude and phase shift of a complex number in rectangular form
given: z  a  jb

b
z  a 2  b2   tan 1  
a 
lesson15et438a.pptx

2
 10/28/2015

Computing Transfer Function Values

5

Example 15-1: A self-regulating tank has a transfer function of the form shown
below.
H(s) G

Q(s) 1    s

The tank has a time constant, =1590 seconds and a gain, G=2000 s/m2.
Determine the amplitude and phase shift of the system to a sinusoidal flow input
of 0.0001592 Hz

Solution: Substitute values of G, , and jw into transfer function and compute the
gain magnitude and phase shift.

lesson15et438a.pptx

Example 15-1 Solution (1)

6

G  2000 s/m 2   1590 s f  0.0001592 Hz

H( jw) 2000

Q( jw) 1  1590  jw

w  2p  f w  2p  0.0001592 Hz  0.001 rad/s

Place radian frequency into transfer function and compute complex number
H( j0.001) 2000 2000
   566.9  j 901
Q( j0.001) 1  1590  j0.001 1  j1.590

Convert the result to polar form to find magnitude and phase shift

lesson15et438a.pptx

3
 10/28/2015

Example 15-1 Solution (2)

7

Complex 2000 2000

G( j0.001)    566.9  j901
gain 1  1590  j0.001 1  j1.590

b
G(jw)  a 2  b 2   tan -1  
a 
Magnitude Calculation Phase shift
a  566.9 b  -901
b   901 
  tan -1    tan -1    58

Ans
G(j0.001)  566.9 2   901 a   566.9 
2

G(j0.001)  1065
At 0.001 rad/sec, the system has a gain
of 60.5 dB and the output changes in height
dB  20  log(1065)  60.5 dB Ans lag the flow changes by 58 degrees

lesson15et438a.pptx

Constructing Bode Plots Using MatLAB

8

MatLAB has control system toolbox functions for defining Linear Time-invariant
systems (LTI) and constructing the Bode plots.

Use tf and bode functions to create LTI and plot. Introducing zpk function

sys = zpk(z,p,k) Turns arrays of zeros, poles and gains into LTI called sys

Where z = array of transfer function zeros

p = array of transfer function poles
k = array of transfer function gains

lesson15et438a.pptx

4
 10/28/2015

Constructing Bode Plots Using MatLAB

9

Example 15-2: Construct the Bode plot for the given transfer function shown in
factored form using MatLAB control toolbox functions.
Vo (s) 0.005  s

Vi (s) 0.001 s  10.001 s  1

Solution: Transfer function has one zero at s=0 and two poles at s=-1/0.001=-
1000

Dividing the transfer function denominator and numerator by 0.001 places it

into standard form
 1   0.005 
0.005  s   s
Vo (s)
  0.001   0.001  
5s
 1   0.001 1  0.001 1  s  1000s  1000
Vi (s)
0.001 s  10.001 s  1    s    s  
 0.001  0.001 0.001  0.001 0.001 
lesson15et438a.pptx

Constructing Bode Plots Using MatLAB

10

MatLAB code to produce Bode plot of given transfer function Magnitude

Enter at command prompt of into m-file plot

k= [5] Phase
p=[1000 1000] plot
Bode Diagram

z=[0]
-50

-60

sys=zpk(z,p,k)
Magnitude (dB)

-70

bode(sys)
w=1000
-80

-90

-100
-90

-135
Phase (deg)

-180

-225

-270
1 2 3 4 5
10 10 10 10 10
Frequency (rad/s)

lesson15et438a.pptx

5
 10/28/2015

Constructing Bode Plots Using MatLAB

11

The bode(sys) function can plot more than one transfer function on the same figure
axis. The figure produced by the bode(sys) function can be copied and pasted
into wordprocessors and other programs. To plot more than one transfer function
use the following syntax: bode(sys1,sys2,…).

Example 15-3: Compare the Bode plots of the transfer function given in Example
15-2 to the function given below. Use MatLAB to generate the Bode plots on a
single set of axis. Note: the only difference is the gain is increased by a factor of
100.
Vo (s) 500  s

Vi (s) s  1000s  1000

lesson15et438a.pptx

Constructing Bode Plots Using MatLAB

12

Enter at command prompt of into m-file sys2

k= [5] 0
Bode Diagram

p=[1000 1000] -20

sys1
z=[0]
Magnitude (dB)

-40
sys1=zpk(z,p,k)
k1=[500]
-60

sys2=zpk(z,p,k1)
-80

bode(sys1,’ro- -100
-90

’,sys2,’b-’) -135
Phase (deg)

-180

-225

-270

This changes plot

1 2 3 4 5
10 10 10 10 10
Frequency (rad/s)

styles sys1 &

(blue solid line) lesson15et438a.pptx sys2

6
 10/28/2015

ET 438a Automatic Control Systems Technology

13

End Lesson 15: Bode Plots of

Transfer Functions

lesson15et438a.pptx