CONTROL SYSTEMS

Dr. Anuradha M
Department of Electronics and
Communication Engineering
CONTROL SYSTEMS

Stability of Control Systems

Dr. Anuradha M
Department of Electronics and Communication Engineering
 CONTROL SYSTEM
Methods of Stability Analysis

• Time domain methods

• Routh-Hurwitz Criterion
• Root Locus Technique
• Frequency domain methods
• Bode Analysis
• Nyquist Stability Criterion
FREQUENCY DOMAIN METHODS

Stability of Control Systems: Bode Plot

Dr. Anuradha M
Department of Electronics and Communication Engineering
 FREQUENCY RESPONSE
Advantages

• Frequency response tests are generally simple and can be determined accurately
by use of sinusoidal signal generators and precise measurement equipment.
• Transfer function of complicated systems can be determined experimentally by
frequency response tests
• System may be designed so that effects of undesirable noise are negligible and
such analysis and design can be extended to non linear systems
 FREQUENCY RESPONSE
Advantages

➢ Modeling transfer functions from physical data.

➢ Designing compensators to meet steady state error and transient response

requirements.

➢ Finding stability of systems

FREQUENCY RESPONSE
Sinusoidal Transfer Function

Y(s)=G(s)R(s)
r(t)=Asinωt -> reference input
𝐴ω
R(s)= 2 2
𝑠 +ω

𝐴ω 𝐴ω 𝐴1 𝐴2
Y(s)=G(s)R(s)=G(s) =G(s) = +
𝑠 2 +ω2 (𝑠+𝑗ω)(𝑠−𝑗ω) 𝑠+𝑗ω 𝑠−𝑗ω

𝐺 𝑠 𝐴ω(𝑠+𝑗ω) 𝐺 −𝑗ω 𝐴ω 𝐴𝐺(−𝑗ω)

A1= 𝑎𝑡 𝑠 = −𝑗ω = =
(𝑠+𝑗ω)(𝑠−𝑗ω) −2𝑗ω −2𝑗
FREQUENCY RESPONSE
Sinusoidal Transfer Function

𝐺 𝑠 𝐴ω(𝑠−𝑗ω) 𝐺 𝑗ω 𝐴ω 𝐴𝐺(𝑗ω)
A2= 𝑎𝑡 𝑠 = 𝑗ω = =
(𝑠+𝑗ω)(𝑠−𝑗ω) 2𝑗ω 2𝑗

1 −𝐴𝐺(−𝑗ω) 𝐴𝐺(𝑗ω)
Y(s)= [ + ]
2𝑗 𝑠+𝑗ω 𝑠−𝑗ω

G(jω) = |G(jω)|𝑒 𝑗φ φ = ∠𝐺(𝑗ω)

G(-jω) = |G(jω )|𝑒 −𝑗φ

𝑨|𝑮(𝒋𝝎)| −𝒆−𝒋𝝋 𝒆𝒋𝝋

Y(s)= [ + ]
𝟐𝒋 𝒔+𝒋𝝎 𝒔−𝒋𝝎

ILT (Inverse Laplace Transform):

𝐴|𝐺(𝑗ω)|
y(t) = [−𝑒 −𝑗φ . 𝑒 −𝑗ω𝑡 + 𝑒 𝑗φ . 𝑒 𝑗ω𝑡 ]
2𝑗

𝑨|𝑮(𝒋𝝎)| −𝒋 𝝋+𝝎𝒕
=> y(t)= [𝒆 + 𝒆𝒋(𝝋+𝝎𝒕) ]
𝟐𝒋
 FREQUENCY RESPONSE
Sinusoidal Transfer Function

• Sinusoidal transfer function: Transfer function as a function of jω

• For a linear system sinusoidal input results in sinusoidal output
• The steady state output will have same frequency as the input
𝒓 𝒕 = 𝑨𝒔𝒊𝒏ω𝒕
• Amplitude and phase of output will vary with frequency 𝒚 𝒕 = 𝑨|𝑮(𝒋𝝎)|𝐬𝐢𝐧(𝝋 + 𝝎𝒕)
 FREQUENCY RESPONSE
Sinusoidal Transfer Function

• In the steady state, sinusoidal inputs to a linear system generate

sinusoidal responses of the same frequency, but Output differs in
amplitudes and phase angle from the input

• Differences are function of frequency.

• Frequency response: Steady state response (output) of a system to

sinusoidal inputs of varying frequency.

• It is plot of magnitude and phase as a function of frequency (as only

magnitude and phase change with change in frequency).
https://eng.libretexts.org/Bookshelves/Industrial_and_Systems_Engineering/Introduction_to_Control_Systems_(Iqbal)
/02%3A_Transfer_Function_Models/2.05%3A_Sinusoidal_Response_of_a_System
 FREQUENCY RESPONSE
Sinusoidal Transfer Function

• On a logarithmic scale, the variable is multiplied by a given factor for equal

increments of length along the axis.
• Decade Change –log10
• Octave change – log2
FREQUENCY RESPONSE
Sinusoidal Transfer Function 𝒚 𝒕 = 𝑨|𝑮(𝒋𝝎)|𝐬𝐢𝐧(𝝋 + 𝝎𝒕)

Ex: For the sinusoidal transfer function G(s)=1/(s+1), find

the magnitude and phase of G(s).
1
𝐺 𝑗ω =
ω2 + 1
Procedure:
1. Put s=jω ⇒ 𝑀𝑎𝑔𝑛𝑖𝑡𝑢𝑑𝑒 𝑖𝑛 𝑑𝐵
2. G(jω)=1/(jω+1) 20 log10 𝐺 𝑗ω = 20 log10 1 − 20 log10 ω2 + 1

(𝑠𝑖𝑛𝑐𝑒 log10 1 = 0)
⇒ 𝑀 = −20 log10 ω2 + 1
−1
ω
φ = ∠𝐺 𝑗ω = 0 − tan
1
CONTROL SYSTEMS

Stability of Control Systems: Bode Plot

Dr. Anuradha M
Department of Electronics and Communication Engineering
 FREQUENCY RESPONSE
Bode Plot

• Bode Plot:
Magnitude plot - magnitude vs frequency
Phase plot – phase vs frequency
Open loop transfer function
• Plotted in log scale(to cover large range)
Magnitude in decibels(dB)
Frequency in rad/sec (on log scale)

https://www.electrical4u.com/bode-plot-gain-margin-phase-margin/#:~:text=Bode%20Stability%20Criterion,-
The%20Bode%20Stability&text=For%20a%20Stable%20System%3A%20Both,equal%20to%20the%20gain%20margin.
 FREQUENCY RESPONSE
Bode Plot

• Consider an open loop transfer function

𝑘 𝑠 + 𝑧1 𝑠 + 𝑧2 … . (𝑠 + 𝑧𝑚 )
𝐺 𝑠 𝐻 𝑠 =
𝑠 + 𝑝1 𝑠 + 𝑝2 … . (𝑠 + 𝑝𝑛 )
• Poles and Zeros are the singularities of this function
• Find the expressions of magnitude(M(ω) in dB) and phase(ϕ(ω)) in degrees
of G(jω)H(jω)
• M and ϕ are plotted against log ω to give the magnitude and phase plots:
Bode plot
• Plotted on a Semilog graph sheet
Frequency axis is logarithmic and magnitude axis is linear
FREQUENCY RESPONSE
Semi Log Sheet
FREQUENCY RESPONSE
Bode Plot

• Since it is a logarithmic plot, the product of terms are converted to additions

• Complicated functions can be visualized in terms of individual standard plots
• Since log scale is used, it is helpful in covering a large range of magnitude and
phase

𝑘 𝑠 + 𝑧1 𝑠 + 𝑧2 … . (𝑠 + 𝑧𝑚 )
𝐺 𝑠 𝐻 𝑠 =
𝑠 + 𝑝1 𝑠 + 𝑝2 … . (𝑠 + 𝑝𝑛 )
 FREQUENCY RESPONSE
Bode Plot

𝐾′ ς𝑚𝑖=1(1 + 𝑇𝑧𝑖𝑗ω)
𝐺 𝑗ω =
ς𝑛𝑗=1(1 + 𝑇𝑝𝑗𝑗ω)
𝐾′ ς𝑚 2 2
𝑖=1 1 + 𝑇𝑧𝑖 ω
𝐺 𝑗ω =
ς𝑛𝑗=1 1 + 𝑇𝑝𝑗 2 ω2
𝑚 𝑛

20 log10 𝐺 𝑗ω = 20 log10 𝐾′ + ෍ 20 log10 1 + 𝑇𝑧𝑖 2 ω2 − ෍ 20 log10 1 + 𝑇𝑝𝑗 2 ω2

𝑖=1 𝑗=1

𝑚 𝑛

∠𝐺 𝑗ω = ෍ tan−1 ω𝑇𝑧𝑖 − ෍ tan−1 ω𝑇𝑝𝑗

𝑖=1 𝑗=1

𝑲′ ς𝒎𝒊=𝟏(𝒔 + 𝒛𝒊)
𝑮 𝒔 = → 𝑷𝒐𝒍𝒆 − 𝒁𝒆𝒓𝒐 𝑭𝒐𝒓𝒎
ς𝒏𝒋=𝟏(𝒔 + 𝒑𝒋)

𝑲′ ς𝒎𝒊=𝟏(𝟏 + 𝑻𝒛𝒊𝒔)
𝑮 𝒔 = → 𝑻𝒊𝒎𝒆 𝑪𝒐𝒏𝒔𝒕𝒂𝒏𝒕 𝑭𝒐𝒓𝒎
ς𝒏𝒋=𝟏(𝟏 + 𝑻𝒑𝒋𝒔)
 BODE PLOT
𝑘 1 + τ𝑠
Bode Magnitude Plots 𝐺 𝑠 =
𝑠 1 + τ𝑠
https://www.geeksforgeeks.org/bode-plots-in-control-system/
Simple Function:
1 1
K, 𝑠 𝑁 , 𝑠𝑁 , 1 + τ𝑠 , 1+τ𝑠

Dr. Anuradha M
Department of Electronics and Communication Engineering
 FREQUENCY RESPONSE
Bode Plot for Individual Singularities

• Constant k: Magnitude plot is a constant with magnitude 20logk dB

𝐾 = 20 log10 |𝐾| , ∠𝐾 = 0°

⇒ 𝐾 = 10, 𝑀𝑑𝐵 = 20𝑑𝐵,

⇒ 𝐾 = 100, 𝑀𝑑𝐵 = 40𝑑𝐵

 FREQUENCY RESPONSE
Bode Plot for Individual Singularities

• Simple zero at origin: 𝑮 𝒔 = 𝒔

• Straight line with slope 20 dB/dec passing through 0dB line at ω =1

𝐺 𝑗ω = 𝑗ω ⇒ 𝐺 𝑗ω = ω ⇒ 20 log10 |𝐺(𝑗ω)| = 20 log10 ω

ω = 1, 𝑀 = 0𝑑𝐵
ω = 10, 𝑀 = 100𝑑𝐵
ω = 100, 𝑀 = 40𝑑𝐵
 FREQUENCY RESPONSE
Bode Plot for Individual Singularities

• Multiple zeros at origin (N zeros):𝑮 𝒔 = 𝒔𝑵

• Straight line with slope 20N dB/dec passing through 0dB line at ω =1

𝐺 𝑗ω = (𝑗ω)𝑁 ⇒ 𝐺 𝑗ω = ω𝑁 , 𝑀 = 20𝑁 log10 ω

𝐴𝑡 𝑁 = 2, 𝑀 = 40 log10 ω
𝑁π
∠𝐺 𝑗ω =
2
 FREQUENCY RESPONSE
Bode Plot for Individual Singularities
1 1 1
𝐺 𝑠 = ⇒ 𝐺 𝑗ω = ⇒ 𝐺 𝑗ω = ,
𝑠 𝑗ω ω
• Simple pole at origin: −1
ω −π
𝑀 = −20𝑁 log10 ω ⇒ φ =0− tan =
0 2
• Straight line with slope -20 dB/dec passing through 0dB line at ω =1
𝑁=1
⇒ −20 𝑑𝐵/𝑑𝑒𝑐

N=1

𝑵=𝟐 N=2
-180°
⇒ −𝟒𝟎𝒅𝑩/𝒅𝒆𝒄

1 −𝑁π
𝐺 𝑠 = ⇒ 𝑀 = −20𝑁 log 10 ω = −40 log 10 ω (𝑓𝑜𝑟 𝑁 = 2) ⇒ φ =
𝑠𝑁 2
 FREQUENCY RESPONSE
Bode Plot for Individual Singularities

• Multiple poles at origin (N Poles):

• Straight line with slope -20N dB/dec passing through 0dB line at ω =1
 FREQUENCY RESPONSE
Bode Plot for Individual Singularities

• Simple/multiple zero (zero on real axis)

• Corner Frequency: Value of ω where the slope changes from 0 dB to

20dB/dec

• Magnitude is 0 dB till corner/critical/break frequency and straight line with

slope of 20 dB/dec or 20N dB/dec beyond corner frequency

• Hence approximate (asymptotic) plot has 2 lines (0dB and 20dB/dec line)
meeting at corner frequency ω=1/T

• Determine the true/actual plot

 FREQUENCY RESPONSE
Bode Plot for Individual Singularities

Zero on the real axis: (1 + 𝜏𝑠)

𝐺 𝑗ω = 1 + 𝑗ωτ ⇒ 𝐺 𝑗ω = 1 + ω2 τ2 , ∠𝐺 𝑗ω = tan−1 ωτ
0 𝑑𝐵, ωτ ≪ 1
20 log 𝐺 𝑗ω = 10 log10 (1 + ω2 τ2 ) = ቐ 3 𝑑𝐵, ωτ = 1
20 log10 ωτ , ωτ ≫ 1

asymptotic
Actual plot

1 1 1
𝑠 + 𝑧𝑖 = 𝑧𝑖 𝑠 + 1 𝑤ℎ𝑒𝑟𝑒 = ⇒ 𝑐𝑜𝑟𝑛𝑒𝑟 𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦
𝑧𝑖 𝑧𝑖 τ
FREQUENCY RESPONSE
Bode Plot for Individual Singularities

1
Pole on the real axis: (1+𝜏𝑠)
FREQUENCY RESPONSE
Bode Plot for Individual Singularities

1
Complex conjugate poles or zeros: (𝑠2 +2𝜁𝜔 2
𝑛 𝑠+𝜔𝑛 )
1 𝜔
Normalized form: 𝐺 𝑗𝜔 = 1−𝑢2 +𝑗2𝜁𝑢
, 𝑤ℎ𝑒𝑟𝑒 𝑢=𝜔
𝑛
FREQUENCY RESPONSE
Bode Plot for Individual Singularities

1
Complex conjugate poles or zeros: (𝑠2 +2𝜁𝜔 2
𝑛 𝑠+𝜔𝑛 )
1 𝜔
Normalized form: 𝐺 𝑗𝜔 = 1−𝑢2 +𝑗2𝜁𝑢
, 𝑤ℎ𝑒𝑟𝑒 𝑢=𝜔
𝑛
FREQUENCY RESPONSE
Bode Plot - Parameters

CE => 1+GH=0
➢We know that GH= -1; Gain= 0 dB, Phase=-1800 GH=-1
|GH|=-1
• Gain Crossover Frequency (GCF) =>20 log |𝐺𝐻| = 0𝑑𝐵
• There must be some frequency at which the Magnitude of G(jω)H(jω) =1
• Frequency at which gain is 0dB or the magnitude plot crosses 0dB line
• Phase crossover frequency(PCF)
Phase =-180°
• There must be some frequency at which the phase of G(jω)H(jω) = -180
• Frequency where angle is -1800 or the phase plot crosses - 1800 line.
FREQUENCY RESPONSE
Bode Plot for Individual Singularities

• Gain Margin: Gain measured w.r.t 0 dB axis at PCF

• Difference between zero and gain at PCF
• Phase Margin: Phase measured w.r.t -1800 at GCF
• Difference between phase and - 1800 at GCF

• Stability: GM and PM should be positive (GCF < PCF)

• Marginal stability: both GM and PM are zero or GM = PM (GCF = PCF)

• Unstable: GM and PM are negative (GCF >PCF)

FREQUENCY RESPONSE
Gain Margin and Phase Margin

• GM:
is +ve: if the intersection of line extended form PCF intersects the Magnitude plot
below 0dB line
is –ve: if the line intersects above the 0dB line
• PM:
is +ve: if the intersection of line extended form GCF intersects the phase plot
above - 1800 line
is –ve: if the line intersects, above the - 1800 line
FREQUENCY RESPONSE
Gain Margin and Phase Margin

• Example
• For the Bode plot shown in Figure, determine
• (a) whether the system is stable, (b) Justify the answer
𝐺𝑎𝑖𝑛
𝐺𝑀 = 0 −
GCF 𝑃𝐶𝐹
= 0 − −17 = 17𝑑𝐵
GM

𝑃𝑀 = 180° + φ
= 180 − 150 = 30°

Since GM, PM is +ve,

system is stable
PM

PCF
FREQUENCY RESPONSE
Gain Margin and Phase Margin
FREQUENCY RESPONSE
Second order closed loop system
Consider second order closed loop system:

Closed loop transfer function

Magnitude 𝑆𝑢𝑏𝑠𝑡𝑖𝑡𝑢𝑡𝑒 𝑠 = 𝑗ω, and find T(jω)

FREQUENCY RESPONSE
Second order closed loop system
Magnitude:

Diff. M w.r.to 𝝎𝟐 and equating to zero yields maximum value 𝑴𝒑

Where the frequency is

 FREQUENCY RESPONSE
Second order closed loop system
Cutoff frequency:
The frequency at which the magnitude of the closed loop frequency
response is -3dB below its zero frequency is called cutoff frequency

Bandwidth:
Bandwidth is the frequency at which the frequency has declined 3dB
from its low frequency value.
1
𝑀=
2
1 1
𝐺 𝑗𝜔 = =
1 − 𝑢2 + 𝑗2𝜁𝑢 2

Solving for u

𝜔𝐵
𝑢= ⇒ 𝜔𝐵 = 𝜔𝑛 [ 1 − 2𝜁 2 + 4𝜁 4 − 4𝜁 2 + 2]
𝜔𝑛

https://www.tutorialspoint.com/control_systems/control_systems_response_second_order.htm
FREQUENCY RESPONSE
Example 2
FREQUENCY RESPONSE
Example 2
FREQUENCY RESPONSE
Example 3
FREQUENCY RESPONSE
Example 3

𝐺𝑀 = 0 − −45
= 45
𝑃𝑀 = ∞
𝑇ℎ𝑒𝑟𝑒𝑓𝑜𝑟𝑒 𝑡ℎ𝑒
𝑠𝑦𝑠𝑡𝑒𝑚 𝑖𝑠 𝒔𝒕𝒂𝒃𝒍𝒆
FREQUENCY RESPONSE
Example 4
FREQUENCY RESPONSE
Example 4

𝐺𝑀 = 33𝑑𝐵
= 33 − 4.4 𝑒𝑟𝑟𝑜𝑟
= 28.6 𝑑𝐵
FREQUENCY RESPONSE
Example 4
FREQUENCY RESPONSE
Example 5
FREQUENCY RESPONSE
Example 5

𝐺𝐶𝐹 = 0.18𝑟𝑎𝑑/𝑠
𝑃𝐶𝐹 = 9 𝑟𝑎𝑑/𝑠
GM=0-(-40)=40dB
Error=8dB
GM=40-8=32dB
PM=180-92=88
FREQUENCY RESPONSE
Example 6 – H.W
10(𝑠+1)
Draw the bode plot for the system 𝐺 𝑠 =
(𝑠+2)(𝑠+5)
FREQUENCY RESPONSE
Example 7 – Manual calculations
10
Find the PCF and GM for the system with 𝐺 𝑠 =
𝑠(1+0.5𝑠)(1+0.1𝑠)
FREQUENCY RESPONSE
Example 7 – Manual calculations

𝟏𝟎
𝑮𝑴𝒅𝑩 = −𝟐𝟎 𝐥𝐨𝐠 𝟏𝟎 |𝝎=𝝎𝒑𝒄
𝝎 𝟏 + 𝟎. 𝟓𝝎 𝟐 𝟏 + 𝟎. 𝟏𝝎 𝟐

= 𝟏. 𝟓𝟕 𝒅𝑩
FREQUENCY RESPONSE
Example 8
10
Find the GCF, PCF, PM and GM for the system with 𝐺 𝑠 =
(𝑠+20)
BODE PLOT

Reverse Bode Analysis

Dr. Anuradha M
Department of Electronics and Communication Engineering
FREQUENCY RESPONSE
Reverse Bode Analysis

• Finding the transfer function for the given Bode plot

• Steps:
1. Initial slope of magnitude plot represents poles or zeros at the origin, if
slope is
• -20dB => 1 pole at origin
• -40dB => 2 poles at origin
• 0dB => No pole
• 20dB => 1 zero at origin
• 40 dB => 2 zeros at origin
 FREQUENCY RESPONSE
Reverse Bode Analysis

Step 2: Observe the shift in the magnitude plot at corner frequency which represents
20logK, from which K can be determined

20 log10 𝐺 𝑗𝜔 𝑎𝑡 𝜔=𝜔1 = 20 log10 𝑋

𝐾 𝐾
Where X can be K or 𝑜𝑟 𝐾𝜔 𝑜𝑟 2 𝑜𝑟 𝐾𝜔2 and so on, depends on the initial slope
𝜔 𝜔

Step 3: Identify the corner frequencies

Step 4: If changes in slope is -20dB/decade, then the factor is a simple pole 1/(1+𝜏s)

Step 5: If changes in slope is 20dB/decade, then the factor is a simple zero (1+𝜏s), where
𝝉= 1/𝝎
FREQUENCY RESPONSE
Reverse Bode Analysis – Example 1

Determine the transfer function for the given bode plot

𝑦2 − 𝑦1
𝑆𝑙𝑜𝑝𝑒 =
log 𝜔2 − log 𝜔1
FREQUENCY RESPONSE
Reverse Bode Analysis – Example 1
1
−20𝑑𝐵 →
• 20 log10 𝐺(𝑗𝜔) = 20 log10 𝐾 − 20 log10 𝜔 1+
𝑗ωൗ
ω1
𝑗ωൗ
• 0𝑑𝐵 = 20 log10 𝐾 => K = 1 +20𝑑𝐵 → 1 + ω2

(1+𝑠/𝜔2 (1+𝑠/𝜔3 ) 𝑗ωൗ

+20𝑑𝐵 → 1 +
• Guess the factors , 𝐺 𝑠 = 𝐾 ω3
(1+𝑠/𝜔1 )(1+𝑠/𝜔4 ) 1
−20𝑑𝐵 →
𝑗ω
1 + ( ൗω4 )
• Where 𝜔1 and 𝜔4 are known but 𝜔2 and 𝜔3 are unknown
• So we need to determine 𝜔2 and 𝜔3
𝑦2 −𝑦1 −20−0
• 𝑆𝑙𝑜𝑝𝑒 = =>−20 =
log 𝜔2 −log 𝜔1 log10 𝜔2 −log10 1

• 𝜔2 =10, similarly, 𝜔3 = 100

log10 ω2 = 1 ⇒ ω2 = 101 = 10
FREQUENCY RESPONSE
Reverse Bode Analysis – Example 1

(1+𝑠/10(1+𝑠/100)
•𝐺 𝑠 = 𝐾
(1+𝑠/1)(1+𝑠/1000)

(1+0.1𝑠)(1+0.01𝑠)
•𝐺 𝑠 =
(1+𝑠)(1+0.001𝑠)
FREQUENCY RESPONSE
Reverse Bode Analysis – Example 2

Determine the transfer function for the given bode plot

1
20 𝑑𝐵 𝑡𝑜 0 𝑑𝐵 ⇒ −20𝑑𝐵 →
𝑗ωൗ
1+ ω2
𝑗ωൗ
0 𝑑𝐵 𝑡𝑜 20 𝑑𝐵 ⇒ +20𝑑𝐵 → 1 + ω1

𝑗ωൗ
−20 𝑑𝐵 𝑡𝑜 0 𝑑𝐵 ⇒ +20𝑑𝐵 → 1 + ω4
1
0 𝑑𝐵 𝑡𝑜 − 20 𝑑𝐵 ⇒ −20𝑑𝐵 →
𝑗ω
1 + ( ൗω3 )

15 − 0 15
20 = ⇒ log10 ω2 =
log10 ω2 − log10 ω1 20
Τ
⇒ ω2 = 1015 20
(1 + 𝑠)(1 + 0.001𝑠) (since ω1 = 1 𝑎𝑛𝑑 𝑙𝑜𝑔1 = 0)
𝐺 𝑠 =
(1 + 0.177𝑠)(1 + 0.0056𝑠)
 FREQUENCY RESPONSE
Reverse Bode Analysis – Example 3
Determine the transfer function of the system whose
approximate plot is shown in figure below.
𝑖𝑛𝑖𝑡𝑖𝑎𝑙 𝑠𝑙𝑜𝑝𝑒 + 20𝑑𝐵
⇒ 𝑧𝑒𝑟𝑜 𝑎𝑡 𝑡ℎ𝑒 𝑜𝑟𝑖𝑔𝑖𝑛
40𝑑𝐵 = 20 log10 𝑥 𝑥 → 𝑘ω1
= 20 log10 𝑘 + 20 log10 ω1
Since ω1 𝑖𝑠 1, 𝑙𝑜𝑔1 = 0
⇒ log10 𝑘 = 2 ⇒ 𝑘 = 102 = 100

1
20 𝑑𝐵 𝑡𝑜 0 𝑑𝐵 ⇒ −20𝑑𝐵 →
𝑗ωൗ
1+ ω1
1 𝒔
0 𝑑𝐵 𝑡𝑜 − 20 𝑑𝐵 ⇒ −20𝑑𝐵 → 𝟏𝟎𝟎𝒔(𝟏 + 𝟏𝟎𝟎𝟎)𝟐
𝑗ω 𝑮 𝒔 = 𝒔 𝒔
1 + ( ൗω2 ) (𝟏 + 𝟏)(𝟏 + 𝟏𝟎)
𝑗ω
−20 𝑑𝐵 𝑡𝑜 20 𝑑𝐵 ⇒ +40 → (1 + )2
ω3
FREQUENCY RESPONSE
Reverse Bode Analysis – Example 3
• Initial slope is +20dB/dec indicating a simple
zero at origin
• This line should have crossed zero dB line at
ω=1
but at ω=1 gain is 40dB
• This indicates a shift of 40dB due to constant
• 20 log10 𝐺(𝑗𝜔) = 20 log10 𝑋
• => magnitude at 𝜔1 = 20 log10 𝐾𝜔1
• => 40 =20 log10 𝑘 + 20 log10 𝜔1
=> 20 log k = 40 or k=100
• Corner frequencies: 1,10,1000
 FREQUENCY RESPONSE
Reverse Bode Analysis – Example 3
• Corner frequencies: 1,10,1000
• At ω=1,slope changes from 20dB/dec to 0dB/dec
indicating that there is a pole with -20dB/dec line
from corner frequency at ω=1 hence canceling the
previous slope
• At ω=10,slope changes from 0dB/dec to -20dB/dec to
indicating that there is a pole with -20dB/dec line
from corner frequency at ω=10
• At ω=1000, slope changes from -20dB/dec to
+20dB/dec indicating that there are two zeros with
+20dB/dec line from this corner frequency(resulting in
+40 dB/dec line) which gives a slope of +20 dB/dec
THANK YOU

Dr. Anuradha M
Department of Electronics and
Communication Engineering
[email protected]
+91 80 2672 1983