MECE 3350 - Control Systems

Lecture 3 – Effect of Pole Locations and Block Diagrams

DC Motor: Review
Finding the relationship between input voltage and output speed

va(t) +

+ vb
vf(t)
-

2
Armature Voltage-Controlled DC Motor: Example
Motor torque for Armature Voltage-Controlled DC motor:
(Assume 𝜃 0 = 𝜃̇ 0 = 𝑖! (0) = 𝑇" = 0; Vb = Kb𝜃;̇ La << Ra)
𝑇 =𝐾 𝐼
𝑚 𝑚 𝑓
(𝑇" =Disturbance torque, va=Armature voltage, vb=Back EMF voltage, TL = TLoad, Tm = Tmotor)
)
Find the Transfer Function *!
𝑇𝑚 = 𝑇𝐿 + 𝑇𝑑
Kirchhoff Voltage Law: 𝑅𝑎𝑖𝑎 + 𝐿𝑎𝑖 𝑎 ! + vb – va = 0
Laplace Transform: 𝑅𝑎𝐼𝑎 + 𝐿𝑎𝑠𝐼𝑎 + Vb – Va = 0
Rearrange: 𝐼𝑎 = (Va – Vb ) / (𝑅𝑎 + 𝐿𝑎𝑠)
Replace Tm: 𝑇𝑚 = Km (Va – Vb )/ (𝑅𝑎 + 𝐿𝑎𝑠)
Replace Vb, 𝓛 Kb𝜽̇ = Kbs𝜽, 𝑻𝒎 = Km (Va – Kbs𝜽) /(𝑹𝒂 + 𝑳𝒂𝒔); La<<Ra:Km (Va – Kbs𝜃) /(𝑅𝑎)
Replace TL: 𝑇𝐿 = ℒ 𝐽𝜃̈ + 𝑏𝜃̇ = 𝐽𝑠 " 𝜃 + 𝑏𝑠𝜃;
Equate Tm and TL: Km (Va – Kbs𝜃) / 𝑅𝑎 = 𝐽𝑠 " 𝜃 + 𝑏𝑠𝜃
Rearrange: KmVa/𝑅𝑎 = 𝐽𝑠 " 𝜃 + 𝑏𝑠𝜃 + (KmKbs𝜃/𝑅𝑎) = s(𝐽𝑠 + 𝑏 + KmKb/𝑅𝑎)𝜃
#
Rearrange: = Km / 𝑠[𝑅𝑎 (𝐽𝑠 + 𝑏) + KmKb]
$!

Divide both sides by 𝑹𝒂𝒃 + KmKb to find the time constant 𝑹𝒂 𝑱/(𝑹𝒂𝒃 + KmKb) 3
 Agenda
• Plotting Pole Locations from Transfer Function Denominator

• Effects of Pole Locations on Stability and Time Response

• Converting Transfer Functions to Block Diagrams

• Converting Closed Loop to Open Loop Block Diagrams

4
Effect of Pole Locations
Low Pass Filter Design
*% 6 6/78
*&
= 789:6 = 9:6/78
1
𝑅𝐶 1
𝑓𝑜𝑟 𝑎 𝑢𝑛𝑖𝑡 𝑠𝑡𝑒𝑝 𝑖𝑛𝑝𝑢𝑡 𝑉1; 𝑉2 𝑠 = ∗
1 𝑠
𝑠+
𝑅𝐶
./
) 𝜏
𝑣2 𝑡 = 1 − 𝑒 - ; 𝑤ℎ𝑒𝑛 𝑡 = 𝜏, 𝑣2 = 0.632 ∗ 𝑣1

𝑡𝑖𝑚𝑒 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡 𝜏 = 𝑅𝐶 =
𝑡𝑖𝑚𝑒 𝑛𝑒𝑒𝑑𝑒𝑑 𝑡𝑜 𝑟𝑒𝑎𝑐ℎ 63.2% 𝑜𝑓 𝑓𝑖𝑛𝑎𝑙 𝑣𝑎𝑙𝑢𝑒 𝑎𝑓𝑡𝑒𝑟 𝑎 𝑠𝑡𝑒𝑝 𝑖𝑛𝑝𝑢𝑡

𝑙𝑎𝑟𝑔𝑒𝑟 𝜏 → 𝑠𝑙𝑜𝑤𝑒𝑟 𝑡𝑟𝑎𝑛𝑠𝑖𝑒𝑛𝑡 𝑟𝑒𝑠𝑝𝑜𝑛𝑠𝑒 → 𝑠𝑡𝑟𝑜𝑛𝑔𝑒𝑟 𝑓𝑖𝑙𝑡𝑒𝑟

)
cutoff frequency fc = *∗,∗-

6
Low Pass Filter Poles
* *
+, ' ()
For a unit step input 𝑉1; 𝑉2 𝑠 = * ∗ ; 𝑣2 𝑡 = 1 − 𝑒 - ; 𝜏 = 𝑅𝐶
%&+, %

)
System poles where denominator 0; s = 0, s = − ; System is stable if 𝑅𝐶 𝑖𝑠 +0 𝑣𝑒
-

Increasing τ : decreases the exponential 1/τ, slows the time response, moves the system poles right

𝜏 𝜏

left poles, right poles, slower decay,

faster decay more dominant
7
Determining System Poles from Laplace Table

no zeros
no poles
no zeros
one pole at zero
a>0 no zeros
one pole at -a, stable if a > 0

no zeros
complex conjugate poles at +𝝎𝒋 and -𝝎𝒋
one zero at zero
complex conjugate poles at +𝝎𝒋 and -𝝎𝒋
no zeros
n+1 poles at zero
no zeros
complex conjugate poles at −𝒂 ± 𝝎𝒋, stable if a>0
one zero at -a
complex conjugate poles at −𝒂 ± 𝝎𝒋, stable
8 if a>0
 Solving Mass Spring Damper Systems
Initial Conditions:
f(t) f(t) = unit step input, x(0)=x0, x’(0)=x1
Step 1: Free Body Diagram, ∑ 𝐹 = 𝑚𝑥′′
mx’’ + bx’ + kx = f(t)

f(t)
Step 2: Laplace Transform
ms2X(s) – msx0 – mx1+ bsX(s) – bx0 + kX(s) = F(s)
6
X(ms2+bs+k) – msx0 - mx1 – bx0 = 9
Step 3: Put in form: Numerator / (s2 + 2𝜁𝜔f + 𝜔f g)
X(ms2+bs+k)s = ms2x0 + msx1 + bsx0 + 1
X = (ms2x0 + msx1 + bsx0 + 1) / (ms2+bs+k)s
X = (s2x0 + sx1 + sbx0/m + 1/m) / (s2 + sb/m + k/m)s
9
Finding Poles and Zeros
X = (s2x0 + s(x1 + bx0/m) + 1/m) / (s2 + sb/m + k/m)s

Step 4: Write the characteristic equation: (s2 + sb/m + k/m)s = 0

Step 5: Find 𝜔f , 𝜁 , θ = cos-1𝜁 𝑎𝑛𝑑 𝜔o = 𝜔f 1 − 𝜁 g ,
considering the characteristic equation form: s2 + 2𝜁𝜔f + 𝜔f g
+ - -
𝜔* = ,𝜁 = , θ = cos−1( ), 𝜔. = 𝜔* 1 − 𝜁 "
, " +, " +,

Step 6: Find the system poles

(h/i)± (h/i)%kl(m/i)
Poles at s = 0, s = - or s = −𝜁𝜔f ± 𝜔f 𝜁 g − 1
g

(n6 : hn./i)± (n& : hn./i)%kl(n./i)

Step 7: Find the system zeros s = gn.
10
Plotting Poles
Step 8: Plot the poles and zeros on s-plane, interpret damping level

𝑘 1 if m = 1, k=1; how does

𝜔f = 𝜔f ∝ 𝑘 𝜔f ∝
𝑚 𝑚 changing b effect 𝜔f , 𝜁 and poles?
b = 0, 1, 1.5, 2, 2.5, 3, -1, -3
𝑏 1 1
𝜁= 𝜁 ∝𝑏 𝜁 ∝ 𝜁 ∝ 2
2 𝑘𝑚 𝑘 𝑚
1

s = −𝜁𝜔f ± 𝜔f 𝜁 g − 1
-3 -2 -1.5 -1 -0.5 0.5 1 1.5 2

-1

Exercise: plot the poles for changing k, and changing m -2 11

Plotting Poles
Step 8: Plot the poles and zeros on s-plane, interpret damping level

𝑘 1 𝜁 > 1: overdamped, no complex roots

𝜔f = 𝜔f ∝ 𝑘 𝜔f ∝
𝑚 𝑚 𝜁 = 1: critically damped, no complex roots
0 ≤ 𝜁 ≤ 1: underdamped
𝑏 𝜁 = 0: undamped
1 1
𝜁= 𝜁 ∝𝑏 𝜁 ∝ 𝜁 ∝ 𝜁 < 0: unstable
2 𝑘𝑚 𝑘 𝑚

s = −𝜁𝜔f ± 𝜔f 𝜁 g − 1

12
Determining Time Response
𝑘 𝑏 1 1 1
𝜔f = 𝜁= 𝜔1 ∝ 𝑘 and 𝜁 ∝ 𝑏 𝑎𝑛𝑑 and
𝑚 𝑘 𝑚
𝑚 2 𝑘𝑚
Step 9: Find the final value, Plot the shape of the time response
if m=1, k=1; how does changing b effect decay and oscillation?
b = 0, 1, 2, 4, -1, -3

1/k

x0
13
time
Assume x’(0) = 0
Determining Time Response
𝑘 𝑏 1 1 1
𝜔f = 𝜁= 𝜔1 ∝ 𝑘 and 𝜁 ∝ 𝑏 𝑎𝑛𝑑 and
𝑚 𝑘 𝑚
𝑚 2 𝑘𝑚
Step 9: Plot the shape of the time response
if m=1, b=2; how does changing k effect decay and oscillation?
k = 0.5, 1, 2

1/k

x0
14
time
Determining Time Response
𝑘 𝑏 1 1 1
𝜔f = 𝜁= 𝜔1 ∝ 𝑘 and 𝜁 ∝ 𝑏 𝑎𝑛𝑑 and
𝑚 𝑘 𝑚
𝑚 2 𝑘𝑚
Step 9: Plot the shape of the time response
if k=1, b=2; how does changing m effect decay and oscillation?
m = 0.5, 1, 2

1/k

x0
15
time
Questions?
Block Diagrams
Outline

By the end of today’s lecture you should be able to

• Represent a control system using block diagrams

• Simplify block diagrams
• Find the open-loop transfer function of a closed-loop system

18
Applications
Typical Transfer Function

19
Applications
The position control system for a spacecraft platform is governed by the
following equations:

d 2 p(t ) + 2 dp(t ) + 4p(t ) =

θ(t ) r (t ): desired position
dt 2 dt
p(t): current position
v 2 (t) = 8v1(t) v1(t): amplifier input voltage
v1 (t ) = r (t ) − p(t ) v2(t): amplifier output voltage
dθ(t) θ(t): motor shaft position
dt = 0.5v2(t)
How can we represent the system using a block diagram ?
dθ(t)
v 2 (t) = 8v1(t) = 0.5v 2(t) d2 p(t ) + 2dp(t ) + 4p(t ) = θ(t )
dt
dt 2 dt

v1 (t ) = r (t ) − p(t )

20
Block Diagrams
→ Represent the relationship of a system variables graphically.

→ Example: Simplifying the relation between the input voltage and the position of a DC motor

21
Basic Building Blocks

Transfer function Y = X*G(s)

Gain Y = X*𝛼

Z
Sum Z=X+Y

Z
Subtract Z=X-Y

22
Basic Operations
Combining blocks in cascade Multiply

Y = XG(s)H(s)

G(s)H(s)

Moving a pickoff point before a block Multiply all output lines

Y = XH(s)

Y = XH(s)
Y = XH(s) 23
Basic Operations If you are stuck, write a formula starting at the output

Moving a summing point after a block Multiply all input lines

Y = (X+Z)H(s)
Y = XH(s)+ZH(s)
Y = XH(s)+Z ??
H(s) Y = XH(s)+ZH(s)

Moving a summing point before a block Divide summing line

W =W W = Y+Z
W = XH + Z
W = (X+ZA)H
1/H W = XH + ZAH
W = XH + Z(1/H)H
24
Basic Operations
Eliminating a feedback loop Multiply by forward model, Divide by 1 + forward model

𝐻 Y = (X-Y)H
1+𝐻 Y = XH - YH
Y+YH = XH
Y
Y(1+H) = XH
w
Y=X 6:w
What if summing block is positive? Flip the sign!

𝐻
1−𝐻

25
Basic Operations
Eliminating a feedback loop Multiply by forward model, Divide by 1 + forward*back model

𝐻 Y = (X-YG)H
1 + 𝐺𝐻 Y = XH - YGH
Y+YGH = XH
Y(1+GH) = XH
w
Y=X 6:yw
What if summing block is positive? Flip the sign!

𝐻
1 − 𝐺𝐻

26
Block Diagram of an armature-voltage controlled DC motor
V (s) − V m (s)
V (s) = (R + Ls)I(s) + ω(s)k m → I(s) =
R + Ls

27
Example 1
Find the open-loop transfer function of the closed-loop system

8{y
Y = X 6:8{yw
28
Example 23
Find the transfer function Y (s)/ R(s) of the system shown.

Procedure:
→ Simply the block diagram
→ Calculate the closed-loop transfer function

29
Example 23 continued

W H2/G4
Step 1: Move pickoff points out of feedback loops
W B
W = X, Y = XG4
Y = XG4, W = XG4B, B = 1/G4 X G4 Y X G4 Y

Step 2: Simplify Feedback Loops

G3G2
Blue circle: =D Green circle: G 2D / [1 + G 2D(H2/G4)] = E Red circle: G 1E/ (1 + G 1E)
1-H1G3G2

30
Exercise 23 continued
G3G2
Blue circle: =D
1-H1G3G2 H2/G4

Green circle: G 2D / [1 + G 2D(H2/G4)] = E

Red circle: G 1E/ (1 + G 1E)

Step 3: Transfer Function: Y(s) = R(s) (G 1E/ (1 + G 1E))

Step 4: Block Diagram

R(s) G 1E/ (1 + G 1E) Y(s)

3)3*3432
Exercise: Substitute in terms for E and D and simplify to Y = ).34325*63*345*6543)3*3432
31
Questions?