SIGNALS AND SYSTEMS USING MATLAB

Chapter 2 — Continuous-time Systems

Luis F. Chaparro
 System definition and types

• System: mathematical transformation of input signal (or signals) into output signal
(or signals) resulting from idealized model of a physical device or process of interest
• Types:
• Static or dynamic system
• Lumped- or distributed-parameter system
• Passive or active system
• Continuous–time, discrete–time or hybrid system

IN P U T OU T P U T

x(t) y(t) = S[x(t)]

S

Continuous-time system S with input x(t) and output y (t)

2 / 25
 Continuous–time system

x(t) ⇒ y (t) = S[x(t)]

Input Output

Properties
• Linearity
• Time-invariance
• Causality
• Stability

A system S is linear if for inputs x(t) and v (t), and constants α and β, superposition
holds, i.e.,
S[αx(t) + βv (t)] = S[αx(t)] + S[βv (t)]
= αS[x(t)] + βS[v (t)]

3 / 25
Examples
• Biased averager Z t
1
y (t) = x(τ )d τ + B, linear if B = 0
T t−T

• Non–linear systems

(i) y (t) = |x(t)|

(iii) v (t) = x 2(t)
• RLC

resistor v (t) = Ri(t), linear

Z t
1
capacitor vc (t) = i(τ )d τ + vc (0), linear if vc (0) = 0
C 0
Z t
1
inductor iL(t) = v (τ )d τ + iL(0), linear if iL(0) = 0
L 0

4 / 25
Operational amplifier
vo (t)

Vsat
i− A
-
+ +
−∆V
+ vd (t)
v− (t) + i+ ∆V
vo (t)
v+ (t)
− − − −Vsat

Op amp: circuit diagram, and input-output voltage relation

Linear model
A → ∞, Rin → ∞ give
virtual short: i− (t) = i+ (t) = 0, vd (t) = v+ (t) − v− (t) = 0

5 / 25
 Time invariance
System S is time-invariant if
x(t) ⇒ y (t) = S[x(t)]
x(t ∓ τ ) ⇒ y (t ∓ τ ) = S[x(t ± τ )]
Examples
• Vocal system

Nasal cavity

Tongue
Vellum
Nose
Epiglottis
Lips
Vocal chords

Lungs

Unvoiced
eu (t)
LTI system s(t)
ev (t)
Speech
Voiced

6 / 25
• Time–varying system

x(t), y (t) input and output of system defined by

y (t) = f (t)x(t), TV if f (t) not constant
• Amplitude modulation (AM) communication system

y (t) = m(t) cos(Ω0t), LTV

m(t) m̃(t)
× Receiver
y(t)

cos(Ω0 t)

Transmitter

AM modulation: transmitter and receiver

7 / 25
• Frequency modulation (FM) communication system
 Z t 
z(t) = cos Ωc t + m(τ )d τ , m(t) message
−∞

FM system non–linear
scale message γm(t) then output is
 Z t 
cos Ωc t + γ m(τ )d τ 6= γz(t)
−∞
FM system time–varying
delay message m(t − λ) then output is
 Z t 
cos Ωc t + m(τ − λ)d τ 6= z(t − λ)
−∞

8 / 25
 • System represented by linear, constant coefficient differential equation: System S,
with input x(t) and output y (t), represented by
dy (t) d N y (t) dx(t) d M x(t)
a0 y (t) + a1 + ··· + = b0 x(t) + b1 + · · · + bM t≥0
dt dt N dt dt M
is linear time–invariant (LTI) if
• IC are zero
• input x(t) is causal (i.e., zero for
i.e., the system is not initially energized

If IC6= 0, x(t) causal consider superposition

Input
x(t) y(t) = yzs (t) + yzi (t)
LTI
system

Initial
conditions
LTI system with x(t) and IC as inputs

9 / 25
• RL circuit: R = 1, L = 1 and voltage source v (t) = Bu(t)

di(t)
v (t) = i(t) + , t > 0, i(0) = I0
dt
solution i(t) = [I0e −t + B(1 − e −t )]u(t)
IC6= 0: (i) I0 = 1 and B = 1
complete response: i1(t) = [e −t + (1 − e −t )]u(t) = u(t)
zero-state response: i1zs (t) = (1 − e −t )u(t)
zero-input response: i1zi (t) = e −t u(t)
(ii) I0 = 1 and B = 2 (double input)
complete response: i2(t) = (2 − e −t )u(t) 6= 2i1(t)
zero-state response: i2zs (t) = 2(1 − e −t )u(t), doubled
zero-input response: i2zi (t) = e −t u(t), same

10 / 25
 1 v(t)
i1(t)

v(t), i1(t)
0.5

−1 0 1 2 3 4 5

2 2v(t)
1.5 i2(t)

2v(t), i2(t)
1

0.5

0
−1 0 1 2 3 4 5
t (sec)

IC= 0, B = 1, 2, circuit is linear

IC = 0, B = 1 : i1(t) = (1 − e −t )u(t)
IC = 0, B = 2 : i2(t) = 2(1 − e −t )u(t) = 2i1(t)

11 / 25
Time invariance: let v (t) = u(t − 1) and I0 initial condition

i3(t) = I0e −t u(t) + (1 − e −(t−1))u(t − 1)

If I0 = 0 then
i3(t) = (1 − e −(t−1))u(t − 1) = i(t − 1), time-invariant
If I0 = 1 then
i3(t) = e −t u(t) + (1 − e −(t−1))u(t − 1) 6= i(t − 1), time-variant

• Averager

1 t
Z
y (t) = x(τ )d τ, (L)
T t−T
shifted input x(t − λ), then output is
Z t Z t−λ
1 1
x(τ − λ)d τ = x(σ)d σ = y (t − λ), (TI)
T t−T T t−T −λ

12 / 25
 Convolution integral

• Application of LTI
If response of a LTI system to v1(t) is i1(t) the response to v (t) applying LTI is i(t).
v1 (t) i1 (t)
1

0.5

0 1 t 0 1 2 t
v(t) i(t)
1
i1 (t)
0.5

0 1 2 t 0 1 2 3 t
−0.5
−i1 (t − 1)
−1

Application of superposition and time invariance to find the response of a LTI system

13 / 25
• Impulse response of LTI system, h(t), is output of the system corresponding to an
impulse δ(t), and initial conditions of zero
• Convolution integral

δ(t) → h(t) (definition)

δ(t − τ ) → h(t − τ ) (TI)
x(τ )h(t − τ ) → x(τ )h(tZ− τ ) (L)
Z ∞ ∞
x(t) = x(τ )δ(t − τ )d τ → y (t) = x(τ )h(t − τ )d τ (L)
−∞ −∞

Z ∞
y (t) = x(τ )h(t − τ )d τ
Z−∞
∞
= x(t − τ )h(τ )d τ
−∞
= [x ∗ h](t) = [h ∗ x](t)

14 / 25
Example: for averager
Z t
1
y (t) = x(τ )d τ, x(t) input, y (t) output
T t−T
1 t
Z
impulse response h(t) = δ(τ )d τ
T t−T

1/T 0 < t < T
=
0 otherwise
Z t
1
ramp response ρ(t) = σu(σ)d σ
T
 t−T
0 t<0
= t 2/(2T ) 0 ≤ t < T
t − T /2 t ≥ T


Note that
d 2ρ(t)
= h(t)
dt 2

15 / 25
Impulse response h(t), unit-step response s(t), and ramp response ρ(t) are related by

 ds(t)/dt
h(t) =
d 2 ρ(t)/dt 2


Interconnection of systems
y(t)
h1 (t) h2 (t) (a)
x(t)

h1 (t)
y(t)
+ (b)
x(t)

h2 (t)

x(t) e(t)
h1 (t) y(t)
+ (c)
−

h2 (t)

Block diagrams of the connection of two LTI systems with impulse responses h1 (t) and h2 (t) in (a)
cascade, (b) parallel, and (c) negative feedback

16 / 25
Cascade
y (t) = [[x ∗ h1] ∗ h2](t) = [x ∗ [h1 ∗ h2]](t) = [x ∗ [h2 ∗ h1]](t), (commute)
Parallel
y (t) = [x ∗ h1](t) + [x ∗ h2](t) = [x ∗ (h1 + h2)](t)
Negative feedback
y (t) = [h1 ∗ e](t)

error signal e(t) = x(t) − [y ∗ h2](t)

closed loop impulse response h(t) = [h1 − h ∗ h1 ∗ h2](t), (implicit)

17 / 25
Example: cascading of two LTI systems
1
∆
u(t) + x(t) y(t)
× Averager
−

Delay ∆

Equivalent block diagram

1
∆
+ y(t)
u(t) s(t)
Averager ×
−
Delay ∆


1 t
Z 0 t<0
s(t) = u(τ )dτ = t/T 0 ≤ t < T
T t−T
1 t≥T

1
y (t) = [s(t) − s(t − ∆)] approximate impulse response of averager
∆

18 / 25
Example: negative feedback
C
R
- -
+ +
+ + +
vi (t)
vo (t) = vi (t) _ vo (t)
+
vi (t) _
− −

(a) (b)
R

R1
-
+
v1 (t) R2 +
+
_
+
v2 (t) vo (t)
_
−

Operational amplifier circuits: (a) voltage follower, (b) inverting integrator, and (c) adder with
inversion

19 / 25
 Causality

• Cause and effect relation between input and output

• For τ > 0, when considering causality let
• time t be the present
• times t − τ be the past, and
• times t + τ be the future
• System S is causal if
• x(t) = 0, IC= 0, output y (t) = 0,
• output y (t) does not depend on future inputs
• LTI system S represented by its impulse response h(t) is causal if

h(t) = 0 for t < 0

output of causal LTI system for causal input x(t) = 0, t < 0
Z t
y (t) = x(τ )h(t − τ )d τ
0

20 / 25
 Graphical computation of convolution

S is LTI and causal, h(t) = 0, t < 0, input is causal, x(t) = 0, t < 0, output
Z t Z t
y (t) = x(τ )h(t − τ )d τ = h(τ )x(t − τ )d τ
0 0

Graphical procedure
• Choose time t0 to compute y (t0 ),
• Plot as functions of τ , x(τ ) and the reflected and delayed t0 , h(t0 − τ ),
• Obtain x(τ )h(t0 − τ ) and integrate it from 0 to t0 to obtain y (t0 ).
• Increase t0 , move from −∞ to ∞

Equal results obtained if x(t − τ ) and h(τ ) used

21 / 25
Example: Unit-step response y (t) of averager with impulse response
h(t) = u(t) − u(t − 1)
h(t − τ )

1 y(t)

τ
t−1 t 0 1

x(τ )
0 1 t
1

τ
0

22 / 25
Example: Graphical computation of the convolution integral when
x(t) = h(t) = u(t) − u(t − 1)
h(t − τ )

1 y(t)

τ
t−1 t 0 1

x(τ )
t
0 1 2
1

τ
0 1

23 / 25
 BIBO stability

• Bounded-input-bounded-output (BIBO) stability: for a bounded x(t) the output y (t)

is also bounded
• LTI S is BIBO stable if
Z ∞
|h(t)|dt < ∞, (absolutely integrable)
−∞
Indeed
∞ ∞
Z  Z
 
|y (t)| =  x(t − τ )h(τ )dτ  ≤ M |h(τ )|dτ ≤ MK < ∞
−∞ −∞

Example: RL circuit (R=L=1)

di(t)
vs (t) = i(t) +
dt
vZs (t) = δ(t), i(0) = 0, i(t) = h(t) = e −t u(t)
∞
|h(t)|dt = −e −t |∞
t=0 = 1
−∞

24 / 25
Example: Positive feedback system (not BIBO stable)
x(t) y(t)
+

Delay τ ×
βy(t − τ )

β
y (t) = x(t) + βy (t − τ )
= x(t) + β [x(t − τ ) + βy (t − 2τ )]
| {z }
y (t−τ )
···
= x(t) + βx(t − τ ) + β 2x(t − 2τ ) + β 3x(t − 3τ ) + · · ·
If x(t) = u(t), β = 2, then
y (t) = u(t) + 2u(t − 1) + 4u(t − 2) + 8u(t − 3) + · · · → ∞

25 / 25