Signal & Linear system

Chapter 2 Time Domain Analysis of CT

System
Basil Hamed
2.1 Introduction
Systems described by Linear, Constant-Coefficient
Differential Equations are Continuous-Time,
Linear Time-Invariant (LTI) Systems

Differential equations like this are LTI

• coefficients (a’s & b’s) are constants ⇒TI

• No nonlinear terms ⇒Linear

Basil Hamed 2
 2.1 Introduction
Many physical systems are modeled w/ Differential Eqs–
Because physics shows that electrical (& mechanical!)
components often have “V-I Rules” that depend on
derivatives

This is what it means to “solve” a differential equation!!

However, engineers use Other Math Models to help solve and analyze differential
eqs–The concept of “Frequency Response” and the related concept of “Transfer
Function” are the most widely used such math models>“Fourier Transform” is
the math tool underlying Frequency Response–Another helpful math model is
called “Convolution”
Basil Hamed 3
2.1 Introduction
Relationships Between System Models
• These 4 models all are equivalent
• …but one or another may be easier to apply to a given
problem

Basil Hamed 4
2.1 Introduction
Example System: RC Circuit
System You’ve seen in Circuits Class that R, L, C circuits are
modeled by Differential Equations:
• From Physical Circuit…get schematic
• From Schematic write circuit equations…get Differential
Equation
• Solve Differential Equation for specific input…get specific
output

Basil Hamed 5
 2.1 Introduction
“Schematic View”:

“System View”:
Circuits class showed how to model this physical system mathematically:

Basil Hamed 6
2.2 Continuous-Time Domain Analysis
For a linear system,

The two components are independent of each other Each

component can be computed independently of the other

Basil Hamed 7
2.2 Continuous-Time Domain Analysis

• Zero-input response • Zero-state response

– Response to non-zero
– Response when f(t) = 0
f(t) when system is
– Results from internal relaxed
system conditions only
– A system in zero state
– Independent of f(t) cannot generate any
– For most filtering response for zero input.
applications (e.g. your – Zero state corresponds
stereo system), we want to initial conditions
no zero-input response. being zero.

Basil Hamed 8
2.2 Continuous-Time Domain Analysis
-Consider that the input “starts at t= t0”:(i.e. x(t) = 0 for t< t0)
-Let y(t0) be the output voltage when the input is first applied (initial
condition)
-Then, the solution of the differential equation gives the output as:

Recall :This part is the solution to the “Homogeneous Differential Equation”

1.Set input x(t) = 0
2.Find characteristic polynomial (Here it is λ+ 1/RC)
3.Find all roots of characteristic polynomial: λi (Here there is only one)
4.Form homogeneous solution from linear combination of the exp {λi (t-to)}
5.Find constants that satisfyBasil
theHamed
initial conditions (Here it is y(t0) ) 9
2.2 Continuous-Time Domain Analysis
In this course we focus on finding the zero-state response
(I.C.’s= 0)

Later will look at this general form…It’s called “convolution”

Basil Hamed 10
 2.2 Linear Constant-Coefficient
Differential Equations
General Form:(Nth-order)

Input: x(t)
Output: y(t) Solution of the Differential Equation

“Homogeneous Solution”
Recall: Two parts to the solution
(i) one part due to ICs with zero-input (“zero-input response”)
(ii) one part due to input with zero ICs (“zero-state response”)

Basil Hamed 11
 2.2 Linear Constant-Coefficient
Differential Equations

Then: y(t) = yZI(t) + yZS(t)

(yzs(t) is our focus, so we will often say ICs = 0)

Basil Hamed 12
 2.2 Linear Constant-Coefficient
Differential Equations
So how do we find yZS(t)?
If you examine the zero-state part for all the example solutions of
differential equations we have seen you’ll see that they all look
like this:

This is called “Convolution”

So we need to find out:

Given a differential equation, what is h(t-λ)

Basil Hamed 13
 2.2 Linear Constant-Coefficient
Differential Equations
Ex. Find 𝑦0 𝑡 the zero-input
component of the response, for a
LTI system described by the
following differential equation:
(𝐷2 + 3𝐷 + 2) 𝑦 𝑡 = 𝐷𝑥(𝑡)
when the initial conditions are
𝑦0 0 = 0, 𝑦ሶ 0 0 = −5.
For zero-input response, we want to find the solution to:
o The characteristic equation for this system is therefore:
(𝐷2 + 3𝐷 + 2) 𝑦0 𝑡 = 0
o The characteristic roots are therefore.

o The zero-input response is

14
Basil Hamed
 2.2 Linear Constant-Coefficient
Differential Equations
To find the two unknowns c1 and c2, we use the initial
conditions

o This yields to two simultaneous equations:

o Solving this gives:

o Therefore, the zero-input response of y(t) is given by:

Basil Hamed 15
 2.3 Unit Impulse response h(t)
Impulse response of a system is response of the system to an
input that is a unit impulse
Impulse Response: h(t) is what “comes out” when δ(t) “goes in”

Use h(t) to find the zero-state response of the system for an input

Basil Hamed 16
2.4 System Response to external Input:
Zero-State Response
The Sifting Property of δ(t):

Above eq. shows that the impulse function sifts out

everything except the value of x(t) at t=t0

Basil Hamed 17
 2.4 System Response to external Input:
Zero-State Response
Approximate any input x(t) as a sum of shifted, scaled pulses

Above figure shows that any signal x(t) can be expressed as a

continuum of weighted impulses

Basil Hamed 18
2.4-1 The Convolution Integral

CONVOLUTION
y(t)= x(t) * h (t) = h(t) * x(t)

Basil Hamed 19
Properties of the Convolution:
1. Commutative: x(t) * h(t)= h(t) * x(t)

2. Associative: x(t) * h1(t) * h2(t)= [x(t)*h1(t)]*h2(t)

[x(t)*h2(t)]*h1(t)
3. Distributive:
x(t)*[h1(t)+h2(t)]=[x(t)*h1(t)]+[x(t)*h2(t)]
4. Convolution with an impulse: x(t)*δ(t)= x(t)

5. Convolution with Unit step:

∞ 𝑡
x(t)*u(t)=‫׬‬−∞ 𝑥 𝜏 𝑢 𝑡 − 𝜏 𝑑𝜏 = ‫׬‬−∞ 𝑥 𝜏 𝑑𝜏

Basil Hamed 20
DISTRIBUTIVITY

Basil Hamed 21
Associative

Basil Hamed 22
Examples of Convolution
Find u(t) * u(t)
∞ 𝑡
y(t)=u(t)*u(t)= ‫׬‬−∞ 𝑢 𝜏 𝑢 𝑡 − 𝜏 𝑑𝜏 = ‫׬‬−∞ 𝑢 𝜏 𝑑 𝜏 = 𝑟 𝑡

Ex 2. P 175
𝑥 𝑡 = 𝑒 −𝑡 𝑢 𝑡
ℎ 𝑡 = 𝑒 −2𝑡 𝑢 𝑡
Find y(t)

Basil Hamed 23
Examples of Convolution
y(t)= x(t)* h(t)=
∞ ∞ −𝜏
‫׬‬−∞ 𝑥 𝜏 ℎ 𝑡 − 𝜏 𝑑𝜏 = ‫׬‬−∞ 𝑒 𝑢(𝜏) 𝑒 −2(𝑡−𝜏) 𝑢 𝑡 − 𝜏 𝑑𝜏

Note that u 𝜏 𝑢 𝑡−𝜏 =1 0<𝜏<𝑡

=0 Otherwise

𝑡 𝑡
𝑦 𝑡 = න 𝑒 −𝜏 𝑒 −2(𝑡−𝜏) 𝑑𝜏 = 𝑒 −2𝑡 න 𝑒 𝜏 𝑑𝜏 = 𝑒 −2𝑡 [𝑒 𝑡 − 1]
0 0
−𝑡
𝑦 𝑡 = (𝑒 − 𝑒 −2𝑡 )𝑢(𝑡)

Basil Hamed 24
 Examples of Convolution
Ex. Use direct integration, find the expression for

Solution:

Because both functions are casual, their convolution is zero for t < 0

𝑒 −𝑎𝑡 − 𝑒 −𝑏𝑡
=( ) 𝑢(𝑡)
𝑎−𝑏
Basil Hamed 25
Examples of Convolution
Ex 2.6 P 178 𝑥 𝑡 = 10𝑒 −3𝑡 𝑢 𝑡
ℎ 𝑡 = [2𝑒 −2𝑡 − 𝑒 −𝑡 ]𝑢 𝑡

Find y(t)
Solution: 𝑦 𝑡 =𝑥 𝑡 ∗ℎ 𝑡 =
10𝑒 −3𝑡 𝑢 𝑡 ∗ 2𝑒 −2𝑡 𝑢 𝑡 − 10𝑒 −3𝑡 𝑢 𝑡 ∗ 𝑒 −𝑡 𝑢 𝑡
∞
𝑦 𝑡 = ‫׬‬−∞ 10𝑒 −3𝜏 𝑢 𝜏 2𝑒 −2 𝑡−𝜏 𝑢 𝑡 − 𝜏 𝑑𝜏 −
∞
‫׬‬−∞ 10𝑒 −𝜏 𝑢 𝜏 𝑒 − 𝑡−𝜏 𝑢 𝑡 − 𝜏 𝑑𝜏

Note that u 𝜏 𝑢 𝑡−𝜏 =1 0<𝜏<𝑡

=0 Otherwise
Basil Hamed 26
Examples of Convolution
𝑡 𝑡
y 𝑡 = ‫׬‬0 20𝑒 −3𝜏 𝑒 −2(𝑡−𝜏) 𝑑𝜏 − ‫׬‬0 10𝑒 −3𝜏 𝑒 −(𝑡−𝜏) 𝑑𝜏

𝑦 𝑡 = [20𝑒 −2𝑡 − 15𝑒 −3𝑡 − 5𝑒 −𝑡 ]𝑢(𝑡)

Basil Hamed 27
Examples of Convolution
Ex. A first-order lowpass filter impulse response is given by
ℎ 𝑡 = −𝛿 𝑡 + 2𝑒 −𝑡 𝑢(𝑡)
Find the zero-state response of this filter for the input
𝑒 𝑡 𝑢(−𝑡)
Solution:

Basil Hamed 28
 Examples of Convolution
From convolution table on Page 177

−𝑡 𝑡
−𝑡 𝑡
𝑒 𝑢 𝑡 + 𝑒 𝑢 −𝑡
2𝑒 𝑢 𝑡 ∗ 𝑒 𝑢 −𝑡 = 2 [ ]
2

Basil Hamed 29
Graphical Convolution Operation
Graphical convolution allows us to grasp visually or
mentally the convolution integral’s result. Many signals
have no exact mathematical description, so they can be
described only graphically.

We’ll learn how to perform “Graphical Convolution,” which

is nothing more than steps that help you use graphical
insight to evaluate the convolution integral.

Basil Hamed 30
Steps for Graphical Convolution x(t)*h(t)

1. Re-Write the signals as functions of 𝜏: x(𝜏) and h(𝜏)

2. Flip just one of the signals around t = 0 to get either x(- 𝜏) or h(- 𝜏)
a. It is usually best to flip the signal with shorter duration
b. For notational purposes here: we’ll flip h(𝜏) to get h(- 𝜏)
3. Find Edges of the flipped signal
a. Find the left-hand-edge τ-value of h(- 𝜏)
b. Find the right-hand-edge τ-value of h(- 𝜏)
4. Shift h(- 𝜏) by an arbitrary value of t to get h(t- 𝜏) and get its edges
a. Find the left-hand-edge 𝜏 -value of h(t- 𝜏) as a function of t
b .Find the right-hand-edge 𝜏 -value of h(t- 𝜏) as a function of t

Basil Hamed 31
Steps for Graphical Convolution x(t)*h(t)
5. Find Regions of 𝜏 -Overlap
a. What you are trying to do here is find intervals of t over which
the product x(𝜏) h(t- 𝜏) has a single mathematical form in terms of 𝜏
b. In each region find: Interval of t that makes the identified
overlap happen
c. Working examples is the best way to learn how this is done
6. For Each Region: Form the Product x(𝜏) h(t- 𝜏) and Integrate
a. Form product x(𝜏) h(t- 𝜏)
b. Find the Limits of Integration by finding the interval of 𝜏 over which
the product is nonzero
c. Integrate the product x(𝜏) h(t- 𝜏) over the limits found in b
7. “Assemble” the output from the output time-sections for all the regions
Basil Hamed 32
Graphically Convolve Two Signals

Ex. Convolve these two signals:

Basil Hamed 33
Graphically Convolve Two Signals
Step #1: Write as Function of 𝜏 Usually

Step #2: Flip h (τ) to get h (-τ)

Basil Hamed 34
Graphically Convolve Two Signals
Step #3: Find Edges of Flipped Signal

Edges

Basil Hamed 35
Graphically Convolve Two Signals
Step #4: Shift by t to get h(t-𝜏) & Its Edge

Basil Hamed 36
Graphically Convolve Two Signals
Step #5: Find Regions of τ-Overlap

Basil Hamed 37
Graphically Convolve Two Signals
Step #5: Find Regions of τ-Overlap

Basil Hamed 38
Graphically Convolve Two Signals
Step #5: Find Regions of τ-Overlap

Basil Hamed 39
Graphically Convolve Two Signals
Step #6: Form Product & Integrate For Each Region

Basil Hamed 40
Graphically Convolve Two Signals
Step #6: Form Product & Integrate For Each Region

Basil Hamed 41
Graphically Convolve Two Signals
Step #6: Form Product & Integrate For Each Region

Basil Hamed 42
Graphically Convolve Two Signals
Step #7: “Assemble”Output Signal

Basil Hamed 43
Examples
Ex. 𝑥 𝑡 = 𝑒 2𝑡 𝑢 −𝑡 ℎ 𝑡 =𝑢 𝑡−3 𝐹𝑖𝑛𝑑 𝑦(𝑡)

Solution: y(t)=x(t)* h(t)

Basil Hamed 44
Examples
𝑡−3 1 2𝜏
Region I: 𝑡 < 3 𝑦 𝑡 = ‫׬‬−∞ 1 𝑒 2𝜏 𝑑𝜏 = 2𝑒
2𝑡−6
1 2𝑡−6 −∞
𝑒
𝑦 𝑡 = 𝑒 −𝑒 =
2 2

Region II: 𝑡 > 3 𝑦 𝑡 =

0 2𝜏 1 2𝜏 1 1
‫׬‬−∞ 1 𝑒 𝑑𝜏 = 2 𝑒 = 2 𝑒 0 − 𝑒 −∞ = = 2

𝑒 2𝑡−6
𝑡<3
𝑦 𝑡 = 2
1
𝑡>3
2

Basil Hamed 45
Examples
Ex. given

Find y(t)

Solution x(t) = 1, for 1≤t≤3 h(t) = t+2, for -2≤t≤-1

= 0, elsewhere = 0, elsewhere

Basil Hamed 46
Examples

Basil Hamed 47
Basil Hamed 48
Basil Hamed 49
Basil Hamed 50
Basil Hamed 51
Basil Hamed 52
Midterm Exam 2017
Determine y(t) where x(t) and h(t) is as shown in
the figure below

Basil Hamed 53
Midterm Exam 2017

Basil Hamed 54
Midterm Exam 2018
Determine y(t) where x(t) and h(t) is as shown in
the figure below

Basil Hamed 55
Midterm Exam 2018
Solution

y(t) = x(t) * h(t) = h(t+1) - 2 h(t-1) + h(t-2)

Basil Hamed 56
2.6 System Stability
Roughly speaking: A “stable” system is one whose output does
not keep getting bigger and bigger in response to an input that
does not keep getting bigger.
Stability : A system is stable if it results in a bounded output
for any bounded input, i.e. bounded-input/bounded-output.
“Bounded-Input, Bounded-Output”(BIBO) stability:
A system is said to be BIBO stable if
for any bounded input: |x(t)| ≤C1< ∞
the output remains bounded: |y(t)|≤C2< ∞

Basil Hamed 57
2.6 System Stability
Ex. Determine if the system is stable or unstable:
i. ℎ1 𝑡 = −3𝑒 2𝑡 𝑢 𝑡
ii. ℎ2 𝑡 = 5𝛿(𝑡 + 5)
iii. ℎ3 𝑡 = 𝑒 4𝑡 𝑢 −𝑡
iv. 𝑦 𝑡 = 𝑒 𝑥 𝑡 u(t)
Solution:
∞
i. ℎ1 𝑡 = ‫׬‬0 −3𝑒 2𝑡 =∞ System is unstable
ii. ℎ2 𝑡 = 5 System is stable
0
iii. ℎ3 𝑡 = ‫׬‬−∞ 𝑒 4𝑡 = 1 System is stable
iv. If x(t)=u(t) 𝑦 𝑡 = Constant System is stable
Basil Hamed 58