Complex Frequency Complex Frequency

Complex Representation of Time-Domain Signals Complex Representation of Time-Domain Signals

Example Example
The s-plane The s-plane

Unit 2: Modeling in the Frequency Domain 1 Complex Frequency

Part 1: Complex Frequency
1 Complex Representation of Time-Domain Signals
Engineering 5821:
Control Systems I 1 Example

Faculty of Engineering & Applied Science

Memorial University of Newfoundland
1 The s-plane

January 7, 2010

ENGI 5821 Unit 2, Part 1: Complex Frequency ENGI 5821 Unit 2, Part 1: Complex Frequency

Complex Frequency
Complex Frequency
Complex Representation of Time-Domain Signals
Example
The s-plane
Consider a typical electrical system: the series RL circuit excited by
a DC source. We can develop a DE for this system:

COVERED ON BOARD
di(t) What if the input was an AC source? The forced response would
L + Ri(t) = Vs be a sinusoid with the same frequency as the input (ω) only it
dt
would differ in amplitude and phase:
The solution for this equation is of the following form:
i(t) = A cos(ωt + θ) + Be αt
i(t) = A + Be αt
Either way, the response consists of constants, exponentials, and
Where A, B, and α are constants determined from the circuit’s sinusoids. We will return to this point shortly...
initial conditions and the DE itself.

Note that the response has two components:

Forced response: A
Natural response: Be αt
ENGI 5821 Unit 2, Part 1: Complex Frequency
Complex Representation of Time-Domain Signals
Complex Frequency
Complex Representation of Time-Domain Signals
Example
The s-plane
Assume we have some time varying quantity called x. We say that
x has a complex frequency s when it can be expressed as follows:

x(t) = <{Xe st }
s=σ
Where X and s are complex numbers and,

s = σ + jω x(t) = <{Xe σt }
= <{X }e σt
and, = Ae σt
X = A + jB or Ce jθ
Notice that s must have units of inverse seconds. Consider the Notice that if σ is negative we have a decaying exponential. If its
signals that can be expressed when s has different values: positive we have a growing exponential (generally a bad thing to
have in a control system).
s=0

x(t) = <{Xe 0t }
= <{A + jB} = A

A constant (i.e. a DC quantity) is represented. ENGI 5821 Unit 2, Part 1: Complex Frequency

Complex Frequency Complex Frequency

Complex Representation of Time-Domain Signals Complex Representation of Time-Domain Signals
Example Example
The s-plane The s-plane

Returning to our example, we can utilize various techniques to Plot of 2 1 − e −2.5t :
arrive at the constants A, B, and α. The final solution is:
Vs  R

i(t) = 1 − e− L t
R
If Vs = 10, R = 5Ω, L = 2H then we have,

i(t) = 2 − 2e −2.5t

This signal is composed of two components with complex

frequencies:
i(t) = <{2e 0t } − <{2e −2.5t }
The values of s are 0 and -2.5.

ENGI 5821 Unit 2, Part 1: Complex Frequency ENGI 5821 Unit 2, Part 1: Complex Frequency
 Complex Frequency Complex Frequency
Complex Representation of Time-Domain Signals Complex Representation of Time-Domain Signals
Example Example
The s-plane The s-plane

s = jω
x(t) = <{Xe jωt } General interpretation: s = σ + jω
= <{Ce jθ e jωt }
x(t) = <{Xe (σ+jω)t }
= <{Ce j(ωt+θ) }
= <{Xe σt e jωt }
We now utilize Euler’s identity:
= e σt <{Xe jωt }
e jθ = cos θ + j sin θ
Continuing the derivation,
On the previous slide we derived <{Xe jωt } = C cos(ωt + θ).
x(t) = <{C cos(ωt + θ) + jC sin(ωt + θ)} Therefore,
= C cos(ωt + θ)
x(t) = Ce σt cos(ωt + θ)
So far these signals with complex frequency s can represent
constants, exponentials, and sinusoids. Thus, the complex This represents a sinusoid with a decaying or growing envelope.
frequency representation covers all types of responses from our RL
circuit.
ENGI 5821 Unit 2, Part 1: Complex Frequency ENGI 5821 Unit 2, Part 1: Complex Frequency

e.g. e −0.4t cos(3x), s = −0.4 + j3 Notice that e −0.4t cos(3x) can be represented by s = −0.4 ± j3.
Why? Since cos is an even function the sign of ω makes no
difference.

Another reason... Let z be a complex number.

1
<{z} = (z + z ∗ )
2
where z ∗ = <{z} − ={z}. Thus we can express a damped or
e.g. e 0.2t cos(4x), s = 0.2 + j4 growing sinusoid as half the sum of complex conjugates.

x(t) = Ce σt cos(ωt)
x(t) = <{Xe (σ+jω)t }
X (σ+jω)t X (σ−jω)t
= e + e
2 2
If we applied Euler’s identity we would see that the complex parts
cancelled out. No matter the sign of ω we get the same result.
 The s-plane
Complex Frequency
Complex Representation of Time-Domain Signals
Example
The s-plane
The s-plane represents a complex frequency s as a point in the
Example complex plane. For different values of s we get different
time-domain responses.
e.g. What is the complex frequency for v (t)?
v (t) = 100e 20t sin(400πt + 75o ) jω Pure sinusoids

The two complex frequencies that both correspond to this signal

are s = 20 + j400π and s = 20 − j400π.
Decaying sinusoids Growing sinusoids
e.g. Express v (t) in the form <{Xe st }.
DC

Since cos(x − 90o ) = sin(x),

o
v (t) = <{100e j(−15 ) e (20±j400π)t } σ

However, the angle in the exponent really should be in radians: Decaying exponentials Growing exponentials

v (t) = <{100e −0.261j e (20±j400π)t }

Decaying sinusoids Growing sinusoids
Would the complex frequency change if v (t) was expressed with
cos as opposed to sin?
ENGI 5821 Unit 2, Part 1: Complex Frequency