Prepared and Compiled by Prof. T.K.Ghoshal & Prof. Smita Sadhu Taught by Prof. Smita Sadhu (SS)
Deptt. of Electrical Engg. Jadavpur University
2010
Syllabus
Part I Concepts of Modulating and Sequential Control. Structure of Modulating Control loops. Process Control terminology. Process Instrumentation diagrams. Controller Implementation: Electronic analog, Digital, Pneumatic Controllers. Self-tuning and Multifunction Controllers, Control Valves. Process Actuators: Electrical, Pneumatic, Hydraulic, Valve positioners. Industrial Instrumentation Systems: Components, structure, specification. Self tuning and Adaptive controllers. Supervisory control : Objectives and Implementation.
Part II
Concept of Processes and Units: Process statics, mass and enthalpy balance. Modelling of process dynamics, Modelling of Chemical processes. Single loop control of standard first order process plants. P-I-D control, Controller tuning, Ziegler-Nichlol's method, Frequency domain design. Feed-forward control, Multi-loop and Cascade control, Interaction and decoupling. Non-linear effects in plants and controllers. Simulation of process control systems. * Boiler Drum Level Control. * Discrete Controllers: Selection of sampling intervals, stability analysis. *
Books: 1. Principles and Practice of Automatic Process Control - Smith and Corripio 2. Principles of Process Control - Patranabis 3. Automatic Process Control - Eckmann 4. Process Control Systems - Shinskey 5. Process Systems Analysis and Control - Coughanowr & Koppel 6. Chemical Process Control - Stephanopoulos 7. Process control - Pollard
2010
Formal Definition:
A physico-chemical process involves physical or chemical change of matter, or conversion of energy
Importance:
Above types of processes produce products like fuels, plastics, cement, metals, adhesives, yarns for cloth, foods, beverages, medicines, All are important in our daily life.
A Chemical Plant
2010
2010
Automatic control
Solution:
Refresh your knowledge in Control System
Specially, times response, Bode Plot, Nyquist plot, stability margins
Consult the reference books Solve the exercises in the problem sheet Read lecture notes of current and previous years Apply your mind and common sense
2010
Some of factors have changed in the past decades, but the subject remains exciting.
2010
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
Formalized later by Electrical Engineers & Shinskey Intensive use of elementary control engg. principles
Requires deep appreciation of freq. domain methods.
2010
A Distillation Column
2010
(a)
(b)
Cooling Coils
10
2010
2010 11
12
Disturbance
A physical quantity or property that can be measured and controlled
SP
Controller
CO
FCE/ Actuator
Valve
PV Process
An element which receives signals in one form and converts it to another form
Transducer/ Transmitter
Feedback signal
2010
13
15
Valves
Diaphragm valve
16
Valve positioners are special actuating systems with feedback of valve position
2010
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
17
A Disturbance is external influence but not a load, that somehow influence the variable to be controlled
Often the disturbances change with time. Speed change in the prime mover is a disturbance to a dc generator Pressure fluctuation of inlet steam is a disturbance to a turbine
2010
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
18
3. Level control
Level
2010
19
Transmitter
A Control Loop requires that Process Variable and some loads, be measured and the values transmitted to the controller for appropriate action. A Transmitter
Measures the variable with a sensor, Converts the sensor signal to a standard level which may be used (remotely) by an indicating instrument, a controller or a recorder.
The sensing element, may or may not be physically inside the transmitter.
2010
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
20
10
Transmitter-II
Standard output signal of Transmitters
DC Current Loop (most common) 4 20 mA 0 20 mA DC Voltage 0 10V 05 V 1 5 V Pneumatic (nearly obsolete) 0.2 1.0 Atmospheric
pressure
Transmitters are generally specified with Full Scale (range) Transmitter reading is usually expressed as percentage or per unit of full scale. Example:
With a range of 1250 C and 40% reading, the current output for a 420mA transmitter would be 4+0.4*16=10.4 ma, corresponding to 1250*0.4=500 C
2010
21
Transmitter-III
Preferred Standard transmitter output:
4-20mA (current loop, suppressed zero)
22
11
Transmitter-IV
Transducer Current Loop Signal To Controller DC Power Source
Transmitter
2010
23
Transmitter-V Transmitter-V
DC Power Source Transducer
Transmitter
Transmitter draws a constant 4 ma current for own powering
2010
24
12
Transmitter-VI
2010
25
Signal out
Transmitter-VII
Power In Pressure Taps
Flow Pipe
Connecting Flanges
26
2010
13
Flow
Actuator CO SP FC
FT
PV
2010
27
2010
28
14
Recap-Control Loop
We have studied essential aspects of a control loop
Loop Structure Set point, Process Variable Valve, Actuator, Final Control Element and Valve positioners Transmitter Process Instrumentation diagram
Coming up:
Controller, Control Law, Feedback and Feed forward
2010 29
Controller
Error = SP-PV Transmitter output should match the controller inputs physically
Type (say current loop) Range Bias FF
Controller output and actuator input also should match physically Signal is one sided, i.e. no reversal of signals
Bias is often used to generate some CO with no error
SP CO
Controller
PV
30
15
Industrial Controllers
2010
31
Load PV
32
16
33
Controllability
(in the sense of process control)
A process to be controlled must be controllable The term controllability has a special meaning in process control context.
This is sometimes called input-output controllability
In the simplest SISO case there is only one process variable of interest. For process controllability in the simple case:
There must exist an input variable (usually a flow), which may be manipulated independently. The process variable should change in response to change in the input variable. For a step change in input, the output should either settle at a constant value or keep on increasing (type one higher type)
2010
34
17
35
Controllability-II
(in the sense of process control)
For multi-output systems, there would be more than one (say m number of) process variables of interest. For input-output controllability of multi-output case
There must be at least m number of input variables which can be independently manipulated. It would be possible to form m number of input output pairs,
each with an unique input (that is input variables cannot be repeated) and each of which is controllable.
Note that
There may be more than one scheme of pairing Some pairing scheme can be better than the other.
2010
Response of a process variable to its input pair should be more significant than the response to unpaired input. This is called diagonal dominance.
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
36
18
2010
37
2010
38
19
D1
q = cd (/4) D22 [ 2(p1 - p2) / (1 - d4) ]1/2 Where q=volumetric flow, cd is a constant D2 = orifice, venturi or nozzle inside diameter D1 = upstream and downstream pipe diameter d = D2 / D1 diameter ratio, =density p1 , p2= upstream and downstream pressures
2010
q=cf(P)
39
Venturi Tube
2010
40
20
The flow coefficient cf is strongly dependent on the units chosen for flow and pressure When pressure is expressed as head, density effect gets cancelled. For mass flow rate, q is to be multiplied by density
2010
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
41
q = c f P + P c f P
~ ~ = c f ( 1 + P / P 1) P q 1 P ~ ~ q 2 P
The flow equation is strongly nonlinear For linear, small signal approximation, local slope may be used. (2% change in pressure drop => 1% change in flow)
2010
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
42
21
43
2010
44
22
q = (
){c f ( ) P } + ( ){c f ( ) P }P P
Small signal linear equations can be derived as above ~ ~ ~ The operating point (q , P , ) should be known
2010
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
45
46
23
0.5 0.4 0.3 0.2 0.1 0.0 0.0 0.2 0.4 0.6 0.8 1.0
F/Fmax
X/Xmax
2010
47
Dye injection
q2
1 0 .8 0 .6 0 .4 0 .2 0 0 .2 0 .4 0 .6 0 .8 1 1 .2 1 .4 1 .6 1 .8 2
Dye q2
Input
v
q1
Colourless Fluid flow rate
q1+q2
0 .8 0 .6 0 .4 0 .2 0 0 .2 0 .4 0 .6 0 .8 1 1 .2 1 .4 1 .6 1 .8 2
T=L/v (Delay)
2010
48
24
Dye injection-II
Step Response
q2
1 0.8 0.6 0.4 0.2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Blue: input
time T=L/v
2010 (Delay)
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
49
Dye injection-III
Dye injection is used routinely for colouring subsidized kerosene to be sold in Fair Price Shops. Processes similar to dye injection are fairly common The delay in response for the dye injection and similar systems is called transport delay. Transport delay is a common phenomenon in chemical processes which involve pipelines and fluid transportation.
2010
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
50
25
2010
51
Equations:
t = time h = Heat transfer coefficient watt / m 2 oC [though actually it is f ( , n )] = temperature difference between oven and surroundings = change in temperature difference
For simplicity, heat transfer is assumed to be proportional to temperature H .t = m.c. + A h.t or , H = m.c. d + A h dt d hA H + = or , dt mc mc
2010
52
26
Transfer function
d hA H & + + = H / hA = dt mc mc mc where = is the time const; ss = H ss / hA hA (s ) ss 1 = or H (s ) H ss 1 + s
ss
H Hss t
2010
t
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
53
non-dominant poles
If one carries out an experiment on a real oven, the response could be a bit different. This is because some effects have been neglected in the simple model. For example:
The oven itself would have its own time constant, which can be much more than the time constant of the charge. The temperature measuring device may have its own time constant
When the above effects are modelled properly, one would get a higher order transfer function. Sometimes one time constant may be significantly larger that the other. This time constant is called dominant and the others non-dominant.
2010
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
54
27
non-dominant poles-II
H (s )
(s )
1 H ss (1 + s 1 )(1 + s 2 )(1 + s 3 ) ss
ss
0.5
2.5
55
Q1
Q2 depends on the pressure at the bottomwhich is proportional to the height of the liquid column
Q2
Q2 H ; Q2 = C f H (say)
56
28
In steady state
~ ~ ~ Q2 = Q2 = Q1 ; H = H ~ ~ ~ ~ ~ Q1 = Q2 = C f H H = (Q1 / C f )
2010
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
57
~ ~ [ where = A / C f = AH / Q1 ] ~ ~ H (s ) Q1 1 1 ( H / Q1 ) = = = . Q1 (s ) C f 1 + s A 1 + s 1 + s
2010 58
29
Q1
Q1: Inlet flow rate Q2: Outlet flow rate H: Height of fluid
This time outlet flow rate Is assumed to be proportional to square root of H Closer to empirical result. Pressure square of flow rate
Q2
Q2 H ; Q2 = C f H (say)
2010
59
In steady state
~ ~ ~ Q2 = Q2 = Q1 ; H = H ~ ~ ~ ~ ~ Q1 = Q2 = C f H H = (Q1 / C f ) 2 ~ ~ C f = Q1 / H
2010
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
60
30
~ Q1 = Q1 + q1 ~ Q2 = Q2 + q2 ~ H = H +h dH dh = dt dt
2010
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
61
or , or , or , or , or , or ,
2010
A A
dH ~ ~ + C f ( H + h) g = Q1 + q1 dt
dh h ~ ~ + C f H 1 + ~ g = Q1 + q1 dt H dh h ~ ~ A + C f Hg 1 + ~ = Q1 + q1 dt 2H A dh h g ~ ~ + C f Hg + C f ~ = Q1 + q1 dt 2 H dh h g A + Cf ~ = q1 dt 2 H ~ dh Q1 A + ~ h = q1 dt 2 H
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
62
31
Q1
R H L
Q2 H ; Q2 = C f H (say)
Q2
A = 2bL = 2 R 2 ( R H ) 2 L = 2 2 RH H 2 L
2010
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
64
32
2010
65
66
33
D A
A = R 2 = ( DH / 2 B) 2 = {( / 4)( D / B) 2 }H 2 = H 2 (say)
Q2 H ; Q2 = C f H (say)
2010 67
2010
68
34
69
Linearization
dY dY ~ ~ = f (Y ,V ) ; in steady state, = 0 ; Y = Y ;V = V dt dt ~ ~ f (Y ,V ) = 0 Considering a small perturbation about the nominal point, ~ ~ Y = Y + y ;V = V + v dY dy f ~ ~ 0 f = = f (Y ,V ) + ( |Y =Y ;V =V~ ) y + ( | ~ ~ )v ~ dt dt Y V Y =Y ;V =V
(Taylor series expansion)
Non linear first order differential equations can be linearized at equilibrium points (steady state values)
2010
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
70
35
Linearization-II
In steady state ~ ~ ~ ~ ~ ~ Q2 = Q2 = Q1 ; H = H ; Q1 = Q2 = C f H ~ ~ ~ ~ H = (Q1 / C f ) 2 C f = Q1 / H
dH = (Q1 Q2 ) / A = (Q1 C f H ) / A dt dh = [{ (Q1 / A)}q1 + { ((Q1 / A) C f H / A)}h] |Q , H ~ ~ 1 Q1 dt H
General Case
2010
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
71
Linearization-III
~ ~ For Process D : A = H 2
dh q1 ~ ~ (Q1 /( H 2 ) C f H /( H 2 )}h] |Q , H = ~ +{ 1 dt A H q1 1 ~ ~ ~ ~ = ~ 2 + {2Q1 H 3 C f ( H 3 / 2 )h} |Q , H 1 H H ~ q1 1 ~ ~ 3 Q1 3 / 2 = ~ 2 {2Q1 H ~ ~ ~ }h H H H2 H ~ q1 Q1 = ~2 ~3 h 2H H
~ 2 ( H ) 3 = ~ Q1
72
36
Many chemical processes may exhibit delayed response. The delays could be due to
Transport of fluid over long pipes Contribution of smaller non-dominant time constants (more about it later)
2010
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
73
2010
74
37
1 Amplitude 0.5 0
0.5 Time (s ec )
1.5
1 Amplitude 0.5 0
2010
75
(n ) 2 G(s) = 2 s + 2 n s + (n ) 2
Self Limiting Oscillatory with overshoot and Undershoot
Non-minimum phase zero causes undershoot
G(s) =
2010
(n ) 2 1 Ts / 2 2 1 + Ts / 2 s + 2n s + (n ) 2
76
38
G ( s) =
K s (1 + s )
G(s) =
K s 2 (1 + s )
77
2010
e sT G(s) = (1 + s )
2010
78
39
1 0.5 0
1/(0.1s+1)2
0 0.5 Time (sec) Step Response 1 1.5
1.5
1/(0.1s+1)6
1 Amplitude 0.5 0
1/(0.1s+1)3
0 0.5 Time (sec) Step Response 1 1.5
1/(0.1s+1)7
0 0.5 Time (sec) Step Response 1 1.5
1 Amplitude 0.5 0
1/(0.1s+1)4
0 0.5 Time (sec) Step Response 1 1.5
Amplitude
0.5 0
1/(0.1s+1)8
0 0.5 Time (sec) Step Response 1 1.5
1 Amplitude 0.5
1/(0.1s+1)5
0.5 Time (sec) 1 1.5
Amplitude
0.5 0
1/(0.1s+1)9
0 0.5 Time (sec) 1 1.5
0 2010 0
79
Frequency Response
Revisit
2010
80
40
In this course we would frequently use Bode Plot and occasionally Nyquist plot.
Bode: Magnitude, Phase against freq Nyquist: Phasor plot in complex plane
2010
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
81
100 100
100 82
2010
41
250
100 20 12.65
44.44 30
2010
83
10
10
-1
-90
-91 0 10
10 10 Frequency (rad/sec)
10
2010
84
42
Double Integrator
Gain at unity Freq = K
G ( j ) =
2
K K = 2 180 deg 2 ( j )
Bode Diagram
10
10 Magnitude (abs)
10
10
-1
10
-2
10
10 Frequency (rad/sec)
10
2010
85
K K = 2 1 + j 1 + ( )
Bode Diag ram
Corner frequency 1/
Asymptotic plot
Magnitude (abs)
Actual plot
gcf = (1 / ) K 2 1 K /
10
0
10
10
Fr equenc y (r ad/s ec )
2010
86
43
Bode Diagram
2010
-90
PIC Lecture 2010 Prof. T.K.Ghoshal 100 Prof. Smita Sadhu &
10
87
10
Frequency (rad/sec)
More General TF
Unit negative slope
10 Magnitude (abs)
2
G (s ) =
K (1 + s ) s(1 + .02 s)
10
10
10 Phase (deg)
-1
Zero slope
-45
-90 -1 10
10 10 Frequency (rad/sec)
10
2010
88
44
G (s ) =
60
40
Magnitude (dB)
1st
corner @ 1 r/s
20
-20
-40
-60 0 10
10
10 10 Frequency (rad/sec)
10
2010
89
2010
90
45
K1e sT G p (s ) = 1 + s
Time Constant
Usually chemical processes are not oscillatory by nature Many plants exhibit 1st order-like self regulating response Have transport delay or higher order fast poles which can be approximated as delay Some text books call the delay as lag !! Separate model required for non- self regulating plants
2010
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
91
G p (s ) =
K1
K1e sT 1 + s
0.8
Amplitude
0.6
0.4
0.2
0.5
T
2010
1 Time (sec)
1.5
92
46
u
0.8
u
t
At t2=T+ y=K(1-e-1)=0.67K
0.67K
y y
0.6
0.4
K=
0.26K
0.2
y u
-0.2
0.05
2010
93
Step Response
Gp (s) =
Ke sT 1 + s
0.8
Amplitude
0.6
0.4
0.2
0.5
T
2010
1 Time (sec)
1.5
94
47
G p (s ) =
1 e
0.0964T
= 0.1
1 e
0.511T
= 0.6
= 0.511; T = 0.0427
0.5
2.5
2010
95
G p (s ) =
= 0.5
1 0.9 0.8 0.7 0.6 Amplitude 0.5 0.4 0.3 0.2 0.1 0
1 e
Fairly good approximation of parameters!
System: sys Time (sec): 0.1 Amplitude: 0.0512
= 0.0512
= 0.628
0.5
= 0.535; T = 0.07
96
2010
48
0.6
0.4
0.2
T
-0.2 0 0.05
0.1 0.15 0.2 0.25 time 0.3 0.35 0.4 0.45 0.5
2010
97
Normalized model
If the steady state O/P with 100% input is also defined as 100% O/P, the gain K becomes unity. The normalized process model becomes:
e sT G p (s ) = 1 + s
2010
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
98
49
Let G1 (s ) = e sT G1 ( j ) = e jT =?? G1 ( j ) = e jT =
??
= ??
K G2 ( j ) = 1 + j
??
2010
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
99
100
50
-T
2010
101
Bode Diagram
Magnitude (abs)
10
-T
10
-1
0 -90 Phase (deg) -180 -270 -360 -450 -540 0 50 100 150 200 250 300 350 400 450
2010
Frequency (rad/sec)
102
51
K K = 2 1 + j 1 + ( )
Bode Diag ram
Corner at 1/
Asymptotic plot
Magnitude (abs)
Actual plot
10
0
2010
10
10
10
Fr equenc y (r ad/s Prof. Smita Sadhu PIC Lecture 2010 Prof. T.K.Ghoshal &ec )
103
Bode Diagram
2010
-90
PIC Lecture 2010 Prof. T.K.Ghoshal 100 Prof. Smita Sadhu &
10
104
10
Frequency (rad/sec)
52
G2 ( s ) =
10
1
1 1 + s
Bode Diagram
Bode Diagram
Magnitude (abs)
10
10
10
-1
10
-1
-45
-90 10
0
10
10
Frequency (rad/sec)
2010
105
1/
0 Magnitude (dB) Phase (deg) -5 -10 -15 -20 0 -45 -90 -135 -180 10 20 30 40 50 60 70 80 90 100
2010
106
53
1/
Phase (deg)
10
10 Frequency (rad/sec)
10
2010
107
Comments
Ordinary first order system has 90 degree Phase Margin and infinite gain margin. Such plants are always stable in closed loop with constant gain. Standard process plant is not ordinary because of the delay Delay has got the capability of providing infinite phase lag Delayed 1st order plants are liable to become unstable in closed loop
2010
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
108
54
Bode Diagram
K=3 K=2
Magnitude (abs)
gcf2 gcf3
10
K=1
Unity gain
10
-1
10
-1
10
10
10
10
1/
2010
Frequency (rad/sec)
109
GM=1/|G(jwp)| PM=180deg+/G(jwg)
Bode Diagram
gcf=wg
Magnitude (abs)
0
10
Gain Margin
-1
Phase (deg)
10
Phase Margin
10
1
pcf=wp
10
2010
Frequency (rad/sec)
110
55
Nyquist plot
G1 ( s ) = e sT
G2 ( s ) = 1 1 + s
Resultant plot
2010
111
2010
112
56
Pad Approximation
The delay term is difficult to handle while computing closed loop response. French mathematician Henri Pad (1863-1953), formulated the "best" approximation of a given function by a rational function of specified order. Using the above method the Laplace transform exp(-sT) of delay, may be approximated by transfer functions of any given order. Normally, for manual computation 1st or second order approximations are used. For computer simulation 5th to 10th order approximations may be employed.
2010
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
113
sT
=e
sT / 2
sT / 2
e sT / 2 = sT / 2 e
(sT / 2) 1 sT / 2 +
(sT / 2)
These terms are changed slightly to accommodate the effect of truncation of higher order terms
2010
PIC Lecture 2010 Prof. T.K.Ghoshal & Prof. Smita Sadhu
114
57
Frequency response Unit step response Time Response of process plant with first order Pad appx. for delay
2010 115
Bode Diagram
10
10
-1
0 Phase (deg)
-90
-180 10
-2
10
-1
10
10
10
Frequency (rad/sec)
2010
116
58
0.5
2.5
2010
117
Time Response of standard plant with 1st order Pad apprx. for delay
Step Response 1
G p (s ) =
K 1 sT / 2 1 + s 1 + sT / 2
0.8
0.6
Open Loop
0.4 Amplitude
0.2
-0.2
-0.4
-0.6
2010
118
59
1 sT / 2 + s 2T 2 / 12 1 + sT / 2 + s 2T 2 / 12
Frequency response Unit step response Time Response of process plant with second order Pad appx. for delay
2010 119
Bode Diagram
Magnitude (abs)
10
10
-1
10
10 Frequency (rad/sec)
10
2010
120
60
0.8
0.6 Amplitude
0.4
0.2
-0.2
-0.4
2010
121
Time Response of standard plant with 2nd order Pad appx. for delay
Step Response 1.2
Open Loop
0.8
Amplitude
0.6
0.4
0.2
-0.2
3 Time (sec)
2010
122
61
Pure delay
0.5 First order Second order Third order Fourth order
-0.5
-1
0.2
0.4
0.6
0.8
1 time
1.2
1.4
1.6
1.8
2010
123
Bode Diagram
First order
Phase (deg)
10
-1
10
10
10
Second order
Phase keeps on increasing as the order becomes higher
10
10
Third order
10
10 Frequency (rad/sec)
10
0 Phase (deg)
Fourth order
2010
0 PIC Lecture 2010 10-1 T.K.Ghoshal & Prof. Smita Sadhu Prof. 10
1242
10
Frequency (rad/sec)
62