Control Systems

Unit 3 & Unit 4 — Complete [Link] Notes

Time Response · Frequency Response · Stability Analysis

Unit Topic Page

Unit 3 Time Response Analysis & Specifications 2

Unit 3 First-Order and Second-Order Systems 2

Unit 3 Steady-State Error and Error Constants 4

Unit 3 P, PI and PID Controllers 5

Unit 3 Routh-Hurwitz Stability Criterion 6

Unit 3 Root Locus Technique 8

Unit 4 Frequency Response Analysis 11

Unit 4 Polar Plots 12

Unit 4 Bode Plots — Theory and Numericals 13

Unit 4 Gain Margin and Phase Margin 15

Unit 4 Relative Stability 17

Unit 4 Nyquist Stability Criterion 17

Unit 4 Time–Frequency Domain Correlation 19

UNIT 3 — TIME RESPONSE ANALYSIS

1. Time Response Analysis

The time response of a system is its output as a function of time when a standard input (step, ramp,
impulse) is applied. It has two components:

• Transient response — decays to zero as t → ∞

• Steady-state response — what remains after transients die out

2. Time Domain Specifications

These are the standard performance measures read from a unit step response:

Specification Symbol Definition

Delay time td Time to reach 50% of final value

Rise time tr Time to go from 10% to 90% of final value

Peak time tp Time to reach the first (maximum) peak

Peak overshoot Mp Maximum deviation above final value (%)

Settling time ts Time to stay within ±2% (or ±5%) of final value

Steady-state error ess Difference between desired and actual final value

For a standard second-order system with transfer function:

G(s) = wn^2 / (s^2 + 2*zeta*wn*s + wn^2)

Key formulas:

• Peak time: tp = pi / wd, where wd = wn * sqrt(1 - zeta^2)

• % Overshoot: Mp = exp(-pi*zeta / sqrt(1-zeta^2)) x 100
• Settling time (2%): ts = 4 / (zeta * wn)
• Damped frequency: wd = wn * sqrt(1 - zeta^2)

3. First-Order System Step Response

Transfer function: G(s) = 1/(τs + 1), where τ = time constant.

Step response: c(t) = 1 − e^(−t/τ)

• At t = τ → c(t) = 0.632 (63.2% of final value)

• At t = 4τ → system is ~98% settled

Numerical 1: First-Order System

A first-order system has G(s) = 5/(2s+1). Find the output for unit step, time constant and settling time.

Solution:

• Time constant τ = 2 sec

• DC gain = 5 → final value of output = 5
 • c(t) = 5(1 − e^(−t/2))
• Settling time ts = 4τ = 8 seconds

At t = τ = 2 s: c(2) = 5(1 − e^−1) = 5 × 0.632 = 3.16

4. Standard Second-Order System — Step Response

Numerical 2: Second-Order Specifications
A second-order system has ωn = 5 rad/s, ζ = 0.6. Find all time-domain specifications.

Parameter Formula Calculation Result

Damped freq. ωd ωn√(1−ζ²) 5 × √(1−0.36) = 5 × 0.8 4 rad/s

Peak time tp π / ωd π/4 0.785 s

% Overshoot Mp e^(−πζ/√(1−ζ²))×100 e^(−2.356) × 100 9.48%

Settling time ts 4 / (ζωn) 4 / (0.6 × 5) 1.33 s

Rise time tr (π−cos■¹ζ)/ωd (3.14−0.927)/4 0.554 s

5. Steady-State Error and Error Constants

The steady-state error depends on the system type (number of free integrators) and the input type.

Input type Error constant Formula ess formula

Step (position) Kp lim(s→0) G(s) 1/(1+Kp)

Ramp (velocity) Kv lim(s→0) s·G(s) 1/Kv

Parabola (accel) Ka lim(s→0) s²·G(s) 1/Ka

System type Kp Kv Ka

Type 0 finite 0 0

Type 1 ∞ finite 0

Type 2 ∞ ∞ finite

Numerical 3: Error Constants

G(s) = 10 / [s(s+2)(s+5)]. Find Kp, Kv, Ka and the steady-state error for ramp r(t) = 2t.

Solution: System type = 1 (one free integrator)

• Kp = lim(s→0) G(s) = ∞ → ess (step) = 0

• Kv = lim(s→0) s·G(s) = 10/(2×5) = 1
• For ramp r(t) = 2t: ess = A/Kv = 2/1 = 2
• Ka = 0 → ess (parabola) = ∞

Steady-state error for ramp 2t = 2 (Type 1 system)

6. P, PI and PID Controllers

 Controller Transfer Function Gc(s) Effect on System

P Kp Reduces ess; increases speed; may cause overshoot

PI Kp(1 + 1/Ti·s) Adds integrator; eliminates ess; may slow response

PD Kp(1 + Td·s) Adds zero; reduces overshoot; improves damping

PID Kp(1 + 1/Ti·s + Td·s) Combines all; best performance; sensitive to noise

Effect of each term:

• P term: Amplifies error signal. Kp↑ → ess↓ but overshoot↑

• I term: Integrates error over time. Eliminates ess completely. Can cause instability.
• D term: Rate of change of error. Reduces overshoot, improves damping. Sensitive to noise.

Numerical 4: PI Controller
Plant G(s) = 1/(s+1) with PI controller Gc(s) = 2(1 + 1/s). Find the closed-loop transfer function and
steady-state error for unit step.

Solution:

• Gc(s) = 2(s+1)/s
• Open-loop: G_OL(s) = [2(s+1)/s] × [1/(s+1)] = 2/s
• Closed-loop: T(s) = (2/s) / (1 + 2/s) = 2/(s+2)
• Final value: c(∞) = 1 → ess = 1 − 1 = 0

PI controller eliminates steady-state error completely.

7. Routh-Hurwitz Stability Criterion

A system is stable if all poles of its closed-loop transfer function lie in the left half of the s-plane (negative
real parts).

Routh-Hurwitz gives a systematic algebraic test without finding roots explicitly.

Steps to Form the Routh Array

For characteristic equation: an·s^n + an-1·s^(n-1) + ... + a1·s + a0 = 0

• 1. Arrange coefficients in two rows

• 2. Fill subsequent rows using cross-multiplication formula
• 3. Count sign changes in the first column → number of RHP poles
• 4. Necessary condition: all coefficients must be present and have the same sign

Numerical 5: Check Stability

Characteristic equation: s^4 + 2s^3 + 3s^2 + 4s + 5 = 0

Row Col 1 Col 2 Col 3

s^4 1 3 5

s^3 2 4 0

s^2 1 5 0
 s^1 −6 ■ 0 —

s^0 5 — —

Row calculations:

• s^2 row: b1 = (2×3 − 1×4)/2 = 1; b2 = (2×5 − 1×0)/2 = 5

• s^1 row: c1 = (1×4 − 2×5)/1 = −6
• s^0 row: d1 = (−6×5 − 1×0)/(−6) = 5
First column: 1, 2, 1, −6, 5 → 2 sign changes

UNSTABLE: 2 poles in the Right Half Plane

Numerical 6: Find Range of K for Stability

1 + K/[s(s+2)(s+4)] = 0 → Characteristic eq: s^3 + 6s^2 + 8s + K = 0

Row Col 1 Col 2

s^3 1 8

s^2 6 K

s^1 (48 − K)/6 0

s^0 K —

For stability, all first column elements > 0:

• (48 − K)/6 > 0 → K < 48

•K>0

Range of K for stability: 0 < K < 48

8. Root Locus Technique

The root locus is the plot of closed-loop poles in the s-plane as gain K varies from 0 to ∞. It starts at
open-loop poles (K=0) and ends at open-loop zeros (K=∞).

Rules for Sketching Root Locus

Rule Statement

1. Branches Equal to number of poles n. Start at OL poles, end at OL zeros.

2. Symmetry Always symmetric about the real axis.

3. Real axis Point on real axis is on locus if total OL poles + zeros to its right is odd.

4. Asymptotes Number = n−m. Angles = (2q+1)×180°/(n−m). Centroid = (Σpoles−Σzeros)/(n−m)

5. Breakaway pt Solve dK/ds = 0. Verify that s lies on the root locus.

6. jω crossing Use Routh array → auxiliary equation gives K_critical and crossing frequency ω.

Numerical 7: Complete Root Locus

G(s) = K/[s(s+3)(s+6)] with unity feedback. Find breakaway point and jω crossing.
Step-by-Step Solution
Poles: s = 0, −3, −6 (n=3, m=0)

Asymptotes: n−m = 3

• Angles: (2q+1)×180°/3 = 60°, 180°, 300°

• Centroid: (0−3−6)/3 = −3

Breakaway point:

• K = −s(s+3)(s+6) = −s^3 − 9s^2 − 18s

• dK/ds = −3s^2 − 18s − 18 = 0 → s^2 + 6s + 6 = 0
• s = (−6 ± √12)/2 = −1.27 and −4.73
• s = −1.27 lies between 0 and −3 → valid breakaway point ✓

jω-axis crossing (Routh array for CE: s^3 + 9s^2 + 18s + K = 0):

s^3 1 18

s^2 9 K

s^1 (162−K)/9 0

s^0 K —

• For marginal stability: (162 − K)/9 = 0 → K = 162

• Auxiliary equation: 9s^2 + 162 = 0 → s^2 = −18 → s = ±j4.24 rad/s

Gain K Pole location Stability

K < 162 All in LHP Stable

K = 162 On jω axis at ±j4.24 Marginally stable

K > 162 Two poles in RHP Unstable

Critical gain K = 162 | Crossing frequency ω = 4.24 rad/s

UNIT 4 — FREQUENCY RESPONSE ANALYSIS

1. Introduction to Frequency Response

When an LTI system is driven by a sinusoidal input r(t) = A sin(ωt), the steady-state output is also
sinusoidal at the same frequency but with different magnitude and phase:

c(t) = A|G(jω)| sin(ωt + ∠G(jω))

Replacing s with jω in the transfer function gives the frequency response function G(jω). The two key
quantities are:

• Magnitude ratio (gain): |G(jω)| = output amplitude / input amplitude

• Phase angle: ∠G(jω) = phase shift of output relative to input

2. Frequency Domain Specifications

Specification Symbol Formula Significance

Resonant peak Mr 1/(2ζ√(1−ζ²)) Indicates relative damping; higher Mr = more oscillato

Resonant frequency ωr ωn√(1−2ζ²) Frequency at which Mr occurs

Bandwidth ωB ωn×(...formula) Frequency at −3 dB; higher = faster response

Cut-off rate — Slope of |G(jω)| Ability to reject high-frequency noise

3. Polar Plots (Nyquist Plots)

A polar plot is the locus of G(jω) in the complex plane as ω varies from 0 to ∞. Each point represents a
complex number with magnitude |G(jω)| and phase ∠G(jω).

Numerical 1: Polar Plot of G(s) = 1/(1+sT)

Substituting s = jω: G(jω) = 1/(1 + jωT)

• |G(jω)| = 1/√(1 + ω²T²)

• ∠G(jω) = −tan■¹(ωT)

ω Magnitude Phase

0 1.0 0°

1/T 0.707 −45°

∞ 0 −90°

Locus is a semicircle of radius 0.5, centred at (0.5, 0) in lower half-plane.

Numerical 2: Polar Plot of G(s) = K/[s(1+sT)]

G(jω) = K/[jω(1 + jωT)]

• ω→0: magnitude → ∞, phase → −90°

• ω→∞: magnitude → 0, phase → −180°
 • Real axis crossing (Im = 0): ω = 1/T, crossing point = −KT/2

If −KT/2 < −1, the system is unstable (encircles critical point −1+j0)

4. Bode Plots
A Bode plot consists of two graphs plotted against log-frequency (log ω):

• Magnitude plot: 20 log|G(jω)| in dB vs. log ω

• Phase plot: ∠G(jω) in degrees vs. log ω

Factor Magnitude (dB) Phase

K (gain) Horizontal at 20 log K 0° (K>0) or −180°

(jω)^±N ±20N dB/dec through 0 dB at ω=1 Constant ±90N°

(1+jωT)^±1 0 for ω■1/T; ±20 dB/dec for ω■1/T 0° to ±90°; ±45° at corner ω=1/T

Second-order 0 for ω■ωn; −40 dB/dec for ω■ωn 0° to −180°; −90° at ω=ωn

Numerical 3: Bode Plot for G(s) = 10/[s(1+0.1s)(1+0.01s)]

Convert to standard form: G(jω) = 10/[jω(1+j0.1ω)(1+j0.01ω)]

Corner Frequencies
• ω1 = 1/0.1 = 10 rad/s
• ω2 = 1/0.01 = 100 rad/s
• Initial slope = −20 dB/dec (from integrator 1/jω)

ω range (rad/s) Slope Reason

0.1 → 10 −20 dB/dec Only integrator active

10 → 100 −40 dB/dec Integrator + first lag (ω1=10)

100 → ∞ −60 dB/dec Integrator + both lags

Starting Magnitude
• At ω = 1: |G(j1)| ≈ 10/1 = 10 → 20 log(10) = 20 dB

Phase Values
ω (rad/s) Approx Phase

0.1 ≈ −90°

1 ≈ −96°

10 ≈ −141°

100 ≈ −219°

∞ −270°

5. Gain Margin and Phase Margin

These are the two most important relative stability measures read directly from a Bode plot.

Term Symbol Definition Formula

Phase crossover freq ωpc Frequency at which phase = −180° ∠G(jωpc) = −180°

Gain crossover freq ωgc Frequency at which |G(jω)| = 1 |G(jωgc)| = 0 dB

Gain margin GM How much gain can increase beforeGM(dB) = −20 log|G(jωpc)|
instability

Phase margin PM How much phase can decrease before = 180° + ∠G(jωgc)
PM instability

Stability rules:

• Stable: GM > 0 dB AND PM > 0°

• Unstable: GM < 0 dB OR PM < 0°
• Good design target: GM ≥ 6 dB and PM = 30° to 60°

Numerical 4: GM and PM for G(s) = 50/[s(1+0.1s)(1+0.5s)]

Step 1 — Find ωpc (phase = −180°)

∠G(jω) = −90° − tan■¹(0.1ω) − tan■¹(0.5ω) = −180°

Use: tan■¹(A) + tan■¹(B) = 90° when AB = 1

(0.1ω)(0.5ω) = 1 → 0.05ω² = 1 → ωpc = √20 = 4.47 rad/s

Step 2 — Find |G(jωpc)|

|G(j4.47)| = 50 / [4.47 × √1.2 × √6] = 50 / 11.99 = 4.17

GM = 1/4.17 = 0.24 → GM(dB) = 20 log(0.24) = −12.4 dB → UNSTABLE

Step 3 — Find ωgc and Phase Margin

By trial, ωgc ≈ 12.5 rad/s
∠G(j12.5) = −90° − 51.3° − 80.9° = −222.2°

PM = 180° + (−222.2°) = −42.2° → Negative PM confirms unstable system

6. Relative Stability — Numerical

Numerical 5: Find K for Phase Margin = 30°
G(s) = K/[s(s+1)(s+2)]. Find K so that PM = 30°.

Step 1: For PM = 30°, phase at ωgc must be −150°:

−90° − tan■¹(ω) − tan■¹(ω/2) = −150° → tan■¹(ω) + tan■¹(ω/2) = 60°

Step 2: Solving by trial:

• At ω = 0.8: tan■¹(0.8) + tan■¹(0.4) = 38.7° + 21.8° ≈ 60° ✓

• ωgc ≈ 0.8 rad/s

Step 3: Set |G(j0.8)| = 1:

K / [0.8 × √(1.64) × √(4.64)] = 1 → K = 0.8 × 1.28 × 2.154

 K = 2.21 gives Phase Margin = 30°

7. Nyquist Stability Criterion

The Nyquist criterion uses the complete Nyquist plot (ω from −∞ to +∞) to determine closed-loop stability
from the open-loop frequency response.

Nyquist Stability Theorem

Z = N + P

Symbol Meaning

Z Number of closed-loop poles in RHP (want Z = 0 for stability)

N Number of CLOCKWISE encirclements of the critical point −1+j0

P Number of open-loop poles in RHP

For stability: Z = 0 → N = −P (i.e. N counter-clockwise if P>0, N=0 if P=0)

Visual Interpretation
Case Nyquist plot N Conclusion

Stable Does NOT encircle −1+j0 0 Z = 0, stable ✓

Marginally stable Passes through −1+j0 — On stability boundary

Unstable Encircles −1+j0 CW 2 Z = 2, unstable

Numerical 6: Nyquist Analysis for G(s) = K/[s(s+1)(s+2)]

Open-loop poles: s = 0, −1, −2 → P = 0
For stability: N = 0 (no encirclements of −1+j0)

Real Axis Crossing of Nyquist Plot

Setting Im[G(jω)] = 0: ω = √2

At ω = √2: Re[G(j√2)] = −K/6

K value Crossing point Encirclements Stability

K=1 −1/6 = −0.167 N=0 Stable ✓

K=6 −1.0 N = 0 (boundary) Marginally stable

K = 10 −1.67 N=2 Unstable ✗

Critical gain K_critical = 6 (Nyquist plot passes through −1+j0)

8. Correlation Between Time and Frequency Response

This connects frequency-domain design to time-domain performance specifications.
 Frequency domain Time domain Relationship

Phase margin PM Damping ratio ζ ζ ≈ PM/100 (for PM < 70°)

Bandwidth ωB Rise time tr tr ≈ 1.8/ωB

Resonant peak Mr % Overshoot Higher Mr → higher overshoot

Gain crossover freq ωgc Settling time ts Higher ωgc → faster settling

Mr vs ζ vs %OS Lookup Table

Mr ζ % Overshoot Phase Margin

1.0 ≥0.707 0% >65°

1.15 0.6 9.5% 60°

1.30 0.5 16% 52°

1.73 0.3 37% 30°

→∞ →0 →100% → 0°

Numerical 7: Full Correlation — ωn = 5 rad/s, ζ = 0.5

Parameter Formula Result

Phase margin PM tan■¹[2ζ/(√(√(1+4ζ■)−2ζ²))] ≈ 54°

Gain margin GM For standard 2nd-order system ∞ (phase never hits −180°)

Bandwidth ωB ωn√[(1−2ζ²)+√(4ζ■−4ζ²+2)] 6.19 rad/s

Resonant peak Mr 1/(2ζ√(1−ζ²)) 1.155

Resonant freq ωr ωn√(1−2ζ²) 3.54 rad/s

% Overshoot e^(−πζ/√(1−ζ²)) × 100 16.3%

Cross-check: Mr = 1.155 → ζ ≈ 0.5 → OS ≈ 16% ✓ Consistent!

QUICK REFERENCE SUMMARY

Unit 3 — Key Formulas

Topic Formula

1st-order step response c(t) = 1 − e^(−t/τ) ; at t=τ → 63.2%

Peak time (2nd order) tp = π / ωd

% Overshoot Mp = exp(−πζ/√(1−ζ²)) × 100

Settling time (2%) ts = 4 / (ζωn)

Rise time tr = (π − cos■¹ζ) / ωd

Damped frequency ωd = ωn√(1−ζ²)

ess (step, Type 0) 1/(1+Kp)

ess (ramp, Type 1) 1/Kv

Routh stability Count sign changes in first column = RHP poles

Root locus breakaway dK/ds = 0 (verify on locus)

jω crossing Auxiliary equation from Routh → K_crit and ω

Unit 4 — Key Formulas

Topic Formula

Gain margin (dB) GM = −20 log|G(jωpc)|

Phase margin PM = 180° + ∠G(jωgc)

Stability (Bode) GM > 0 dB AND PM > 0°

Nyquist criterion Z = N + P ; need Z = 0

Resonant peak Mr = 1 / (2ζ√(1−ζ²))

Resonant frequency ωr = ωn√(1−2ζ²)

PM vs damping ζ ≈ PM/100 (rough rule)

BW vs rise time tr ≈ 1.8 / ωB

Polar plot, 1st-order lag Semicircle from (1,0) to origin, radius 0.5

Phase crossover (ωpc) ∠G(jωpc) = −180°

Gain crossover (ωgc) |G(jωgc)| = 1 (0 dB)

Control Systems — Unit 3 & Unit 4 | [Link] Notes