Scribd
Upload
54 views

Ch-5 Time Res Webpage

This document discusses time response analysis of different order systems. It covers: - Zeroth order systems have a constant transfer function. Examples include potentiometers and DC amplifiers. - First order systems' step response is an exponential function with a time constant τ. A smaller τ means a faster response. - Second order systems' responses depend on the damping ratio ζ. Their step responses are derived for different cases of ζ. - Adding poles and zeros to systems can shape their step responses, with the location of addition impacting the response.

Uploaded by

Tushar Gupta
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
54 views

Ch-5 Time Res Webpage

This document discusses time response analysis of different order systems. It covers: - Zeroth order systems have a constant transfer function. Examples include potentiometers and DC amplifiers. - First order systems' step response is an exponential function with a time constant τ. A smaller τ means a faster response. - Second order systems' responses depend on the damping ratio ζ. Their step responses are derived for different cases of ζ. - Adding poles and zeros to systems can shape their step responses, with the location of addition impacting the response.

Uploaded by

Tushar Gupta
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 48

Chapter 5:Time Response Analysis:

Test signals
Impulse Step Ramp Sinusoidal (frequency response) Laplace Transforms
Impulse (t) step u(t) Ramp t u(t) 1 1/s 1/s2

zeroth order system :


C(s) / R(s) = K ; a constant (algebraic equation)

Examples of zeroth order systems :


Potentiometer dc amplifier dc tachogenerator

First order systems:


C(s) / R(s) = K / (1+s ) Step Response C(t) = K ( 1-t / ) e

;for unit step

K : steady state value of the Function : time constant, smaller, faster response.

300

dc/dt

250

t=0 = K=1

1/

200

150

100

50

t1 t2

Time constant : t2 < t1


0

0.5

1.5

2.5

3.5

4.5

t Error e(t) = e-t / ess = 0 If K 1 and input is not unity


then ess ?

Ramp response of first order systems r (t) = t u(t); R (s) = 1 / s2

C(s) K = ; Let K = 1 R(s) 1 + st


c(s) = 1 s (1 + s )
2

1 1 C (s ) = 2 s s 1+ s

c(t) = t - (1 e

-t/

Error e(t) = (1-e-t /) ess =


If K 1 and input is not unity then ess ?

10 9 8 7 6 5 4 3 2 1 0
0 0 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 9 9

dc/dt = 1 e

-t/

Second order systems

n C(s) = 2 2 R ( s ) s + 2 n s + n
2
Second order example position control

For unit step input


R (s ) = 1 s

Applying final value theorem

C(t) t =1.0
Steady state error for step input = 0

position control Example

r +

Kp

KA

KT Ra

L 1 s(Js+B)

sKb

Taking

KP = 0.1 V / rad KA = 10 V / V KT (Nm/A) = Kb (V/rad/sec. ) = 0.5 J = 1 Kg m2 B = 0.5 Nm / rad / sec

L(s) 1 n = 1 rad / sec and = 0. 5 = 2 r (s) s +s +1

s2 + s + 1 =

- 0.5

( s + 0.5) + (
2

-j

In general s2 + 2 n s + n 2 = (s + )2 + n 12 n

s1,2 = n j n 1
= - n j d ;

d = n 1 - 2

d damped freq of oscillations damping factor n undamped natural frequency

= 0 : undamped

0 < < 1: under damped


roots of s are complex conjugate

For = 1 : Critically damped


roots of s are real and repeating at s = -n

> 1 : over damped


roots of s are real and distinct

Unit step response


r (t) = u(t) ; R(s) =1/s

Case 1: =0

2 n C(s) = 2 2 s(s + n )

c ( t ) = 1- cos nt

(pure oscillation)

Case 2: 0 < < 1

C(s) =

2 n

s (s2 + 2 ns + n2)

c (t) = 1 -

1
e
2

- n t

sin ( d t) + e

- n t

cos ( d t)

c (t) = 1 -

- n t 2

sin ( d t + )

Where - n is real part and d is the imaginary part of the roots

n 1 - 2

- n

where cos =

1+
C(t) 1.0

e- nt

1- 2 1e-nt

1-2

Time(t)

Unit step response of 2nd order system for <1

C(t)
Mp

Tolerance band

1.0

tr tp

Time ts

Time response specification

Peak overshoot (Mp): sin (dt) = 0, dt =

dc = 0 dt

Time at peak overshoot ( tp ) = /d At t = tp ,


c(t p ) = 1 e
n .t p 2
1 2

sin ( + )

Mp = e

Settling time (ts):


Assume tolerance band

5%
3 ts = n

0.05 = e- t

n s

l n (0.05) = n t s or

4 t s for 2% tolerance band is n

Rise time (tr):


c (t r ) = 1 = 1 - tr = d

e - n t r 1
2

sin (d t r + )

Delay time (tr):


c (t d ) = 0.5 = 1 e
- n t d 2

sin (d t d + )

So delay time (td) (1.1+.125+.469 2)/n

Second order system response with ramp input n2 s2 (s2 + 2 ns + n2)


- n t

C(s) =

2 e c( t ) = t + sin ( d t + ) n n 1 2

Steady state error of a second order system with ramp input:

E(s) =

1 s2
2

1-

n2 (s2 + 2 ns + n2)

1 s + 2ns = 2. 2 2 s s + 2ns + n
ess

2 =

For the example taken, = 0.5 & n = 1 ess = 1 unit

Error coefficients
R(s) + H (s) E(s) C(s) G(s)

C (s) = R (s) . G (s) / ( 1+G (s) H (s) ) E (s) G (s) = C (s)

E (s) = R (s) (1 + G (s)H(s))


Steady state error (ess ) = Lt s E (s)
s0

Step input R (s) = 1 /s (constant position)


1 1 e ss = Lt s . s 0 s 1 + G (s ) H ( s )
s 0

Lt G (s ) H (S) = K p (position error coeff.)


So

1 e ss = 1+ K p

Ramp input R (s) = 1


2 /s

(constant Velocity)

1 1 e ss = Lt s 2 . s0 s 1 + G (s ) H (s )
s0

Lt sG (s ) H (S) = K v (Velocity error coeff.)


e ss 1 = Kv

So

Parabolic input
R (s) = 1 /s3 (constant Acceleration)

1 1 e ss = Lt s 3 . s0 s 1 + G (s ) H (s )
s 0

Lt s G (s)H(S) = K a (Accleration error coeff.)


2

So

e ss

1 = Ka

(s + z1 )(s + z 2 )(....) G (s ) H (s ) = n s (s + p1 )(s + p 2 )(....) n type of the system

Type 0 system
K p = Lt G (s ) H (s ) = finite value
s0

1 e ss = = finite value 1+ Kp

K v = Lt sG ( s ) H ( s ) = 0 e ss =
s 0

K a = Lt s G ( s ) H ( s ) = 0 e ss =
2 s 0

Type 1 system

K p = Lt G (s) H (s) = e ss = 0
s0

K v = Lt sG (s) H(s) = finite


s0

e ss = 1 K v = finite

K a = Lt s G(s)H(s) = 0 ess =
2 s0

Type 2 system

K p = Lt G (s) H (s) = e ss = 0
s0

K v = Lt sG (s) H (s) = e ss = 0
s0

K a = Lt s G (s ) H (s ) = finite
2 s0

e ss = 1 K a = finite

Input Type

u(t)
1 1+ Kp +

t u(t)

t u(t) 2

0 1 2

1 Kv


1 Ka

0 0

Compensation Techniques:
Example

R +

1
s(s+1)

K C(s) K G= and = 2 s(s + 1) R (s) s + s + K

1 n = K , = 2 K
If K = 100 n = 10 rad / sec, = 0.05

Mp= 85 %

ts (2%) = 4 n = 8sec.
ess for ramp input = 2 /n = 0.01 unit

While the steady state error is acceptable Mp is not. Damping () to be increased to say, 0.5

1 Then 2 n =1 n = =1 2 0.5
ess for ramp input = 2 /n = 1unit
ess increases to 1 - not acceptable. change the amplifier configuration.

Derivative error Compensation


R + 1 K+sKD s(s+1) C

Required Specifications:

= 0.5 and ess for ramp input = 0.01

K + sK D G(s)H(s) = s(s + 1)
C (s ) K + s KD = 2 R (s ) s + s + s K D + K
n = K = 10 rad/sec 2 n = 1 + KD for = 0.5, KD = 9

4 t s ( 2%) = = 0.8 sec. n

ess for ramp input remains the same

Derivative output Compensation

R +

1 s(s+1)
sKt

Introduce another feed back loop

K G (s)H(s) = s(s + 1 + K t ) K C(s) R (s) = 2 s + s (1 + K t ) + K


Required Specifications:

= 0.5 and ess for ramp input = 0.01

K v = Lt sG (s ) H (s )
s 0

K = = 100 - - - - - (1) 1+ Kt

2 n =1+ Kt or K =1+ Kt (2)

K = 100 K = 10000 K t = 99

Comparing both methods derivative output compensation: additional transducer required gain reduced due to feedback so high gain amplifier required and will have a higher n and hence lesser ts

Integral Control
+ KI KA + s
1 s ( s + 1)

K As + K I G (s ) H (s ) = 2 s (s + 1)

type 2

ess

for step input = 0 ess for ramp input = 0 ess for parabolic input is finite

However system becomes of 3 rd order stability problem arises and will be discussed in later chapters

Addition of Zeros and Poles to the system


Zero added to closed loop Transfer function

C( s ) (1+Tzs) = 2 2 R( s ) s +2 s +n n C( s ) Ts = 2 +2 2 2 R( s ) s +2 s +n s +2 s +n n n
2 n 2 n z

2 n

y(t)= y1(t)+ Tz dy1/dt

Unit step response when zero is added to closed loop Transfer function at various mentioned positions

Unit step response showing the effect of zero addition to closed loop Transfer function.

Zero added to forward path


K(1 + Tzs) G(s) = , K = 6, p1 = 1, p 2 = 2 s(s + p1 )(s + p 2 ) C(s) 6(1 + Tzs) = 3 R (s) s + 3s 2 + (2 + 6Tz )s + 6

Unit step response when zero is added to Forward path at various positions

Pole added to forward path


C(s) n = R ( s ) s(s + 2n )(1+ Tps) C(s) n = R ( s ) Tps3 + (1 + 2n Tp )s2 + 2n s + n 2
2 2

Unit step response when pole is added to Forward path at various mentioned positions

Pole added to closed loop Transfer Function

C( s ) n = 2 2 R ( s ) (s + 2ns + n )(1+ Tps)

Unit step response when pole is added to closed loop transfer function at various mentioned positions.

You might also like