EE2010 Signals and Systems (Tutorial #8)

1. Compute the Fourier series coefficients of x (t )

⎛ 5π ⎞ ⎛ π⎞
x (t ) = sin 2 ⎜ t ⎟ + cos ⎜ 3π t + ⎟
⎝ 4 ⎠ ⎝ 3⎠
Solution:
ω1 ( 5π 4 ) 5
First, = = , a rational number ⇒ x (t ) is periodic!
ω2 3π 12
1 1 ⎛ 5π ⎞ ⎛ π⎞ ⎛ 1 − cos 2θ ⎞
x (t ) = − cos ⎜ 2 × t ⎟ + cos ⎜ 3π t + ⎟ ⎜Q sin θ =
2
⎟
2 2 ⎝ 4 ⎠ ⎝ 3⎠ ⎝ 2 ⎠
1 1 ⎛ 5π ⎞ ⎛ π⎞
= − cos ⎜ t ⎟ + cos ⎜ 3π t + ⎟
2 2 ⎝ 2 ⎠ ⎝ 3⎠
⎧ 5π ⎫ π
⇒ ω0 = GCF {ω1 , ω2 } = GCF ⎨ , 3π ⎬ =
⎩ 2 ⎭ 2
⎛ ⎧4 2⎫ 2π 2π π ⎞
⎜ Or, LCM {T1 , T2 } = LCM ⎨ 5 , 3 ⎬ = 4 ⇒ ω0 = T = 4 = 2 ⎟
⎝ ⎩ ⎭ 0 ⎠ 1
 1 1 ⎛ 5π ⎞ ⎛ π⎞
x (t ) = − cos ⎜ t ⎟ + cos ⎜ 3π t + ⎟
2 2 ⎝ 2 ⎠ ⎝ 3⎠
1 1 ⎛ j 52π t 5π
−j t ⎞ 1 ⎛ j ⎜⎛ 3π t + π3 ⎟⎞ − j ⎜ 3π t + ⎟ ⎞
⎛ π⎞
π
= − ⎜e + e 2 ⎟ + ⎜e ⎝ ⎠
+ e ⎝ 3⎠ ⎟ Note: ω0 =
2 4⎝ ⎠ 2 ⎜⎝ ⎟
⎠ 2
⎛π ⎞ ⎛π ⎞ ⎛π ⎞ ⎛π ⎞
1 1 j 5 ⎜⎝ 2 ⎟⎠ t 1 − j 5 ⎜⎝ 2 ⎟⎠ t 1 j π3 j 6 ⎜⎝ 2 ⎟⎠ t 1 − j π3 − j 6 ⎜⎝ 2 ⎟⎠ t
= − e − e + e ⋅e + e ⋅e
2 4 4 2 2
⎧ 1 1 j0 1
⎪ 0 2 = 2 e = 2 ∠0
c =
⎪
⎪ c5 = c−5 = − 1 = 1 e ± j π = 1 ∠ ( ±π )
⎪ 4 4 4
⎪⎪ 1 j π3 1 ⎛ π ⎞ Note: The FS expansion terms
⇒ ⎨ c6 = e = ∠ ⎜ ⎟
⎪ 2 2 ⎝3⎠ must be harmonically related!
⎪ 1 − j π3 1 ⎛ π ⎞
⎪ c−6 = e = ∠⎜ − ⎟
⎪ 2 2 ⎝ 3⎠
⎪ cn = 0, else.
⎪⎩ 2
2. Determine the Fourier series of the sawtooth waveform using
the trigonometric form, and plot its line spectra. Further plot
the truncated (approximated) Fourier series representation of
g(t) using only the dc term and the first 4 ac harmonics terms.
g(t)

… …
Solution:
2π
T0 = 1, ω0 = = 2π , g (t ) = t , for 0 < t < 1
T0
1
1 1 1 1 2 1 1
a0 =
T0 ∫0 g ( t ) dt = ∫0 t dt = t =
2 0 2
(as expected! Q triangle area =
2
×1×1 )

2 1 1
an =
T0 ∫ 0
g (t ) cos n ω0t dt = 2 ∫ t cos 2 n π t dt
0

2 1 1
bn =
T0 ∫
0
g (t ) sin n ω0t dt = 2 ∫ t sin 2 n π t dt
0

For each integration, we need to apply the integration by parts, as follows. 3

 Excursion − Apply the integration by parts: ∫ u dv = u v − ∫ v du
1 1
• ∫ t cos α t dt = α 2
cos α t +
α
t sin α t (in this problem, α = 2 n π )

Proof): Let u = t ⇒ du = dt
1
∫ dv = ∫ cos α t dt ⇒ v=
α
sin α t

1 1 1 1
∫ t cos α t dt = α t sin α t − α ∫ sin α t dt =
α
t sin α t +
α2
cos α t

Similarly,
1 1
• ∫ t sin α t dt = α 2
sin α t −
α
t cos α t (for finding coefficients bn )

Now, back to our problem ...

4
 2 1 1
an =
T0 ∫
0
g (t ) cos n ω0t dt = 2 ∫ t cos 2 n π t dt
0

1
⎡ 1 1 ⎤
= 2⎢ cos 2 n π t + t sin 2 n π t ⎥
⎢⎣ ( 2 π n ) 2π n
2
⎥⎦ 0
⎡ 1 1⎤
1 1
= 2⎢ cos 2 n π t + t sin 2 n π t ⎥ = ⎡⎣ (1 − 1) + ( 0 − 0 )⎤⎦ = 0
⎢ (2π n ) 2
2π n ⎥
⎣ 0 0
⎦
2 1 1
bn = ∫ g (t ) sin n ω0t dt = 2 ∫ t sin 2 n π t dt
T0 0 0

1
⎡ 1 1 ⎤ 2 1
= 2⎢ sin 2 n π t − t c os 2 n π t ⎥ = 0 − [ cos 2 n π ] = −
⎢⎣ ( 2 π n ) π π nπ
2
2 n ⎥⎦ 0 2 n

1 1 ∞ 1
g (t ) = − ∑ sin 2 n π t
2 π n =1 n

5
6
3. Determine the Fourier series of the rectangular pulse train f (t) using the
complex exponential form.
f (t)
A
... ...
T τ τ T0
t
-T0 − 0 − 0 T0
2 2 2 2

Plot the amplitude spectrum and the phase specturm of

the following cases, respectively:
(a). A = 1, T0 = 1, τ = 0.5
(b). A = 1, T0 = 1, τ = 0.25

7
 +∞
2π
Solution: f (t ) = ∑
n =−∞
cn e j n ω0 t , where ω0 =
T0
1 T0 /2 A τ /2 Aτ
c0 =
T0 ∫
−T0 /2
x(t ) dt =
T0 ∫−τ /2
1 dt =
T0
(as expected! Q rectangular pulse area = Aτ )

1 T0 /2 A T0 /2 − j nω0 t
cn =
T0 ∫
−T0 /2
x(t ) e − j n ω0 t dt = ∫−
T0 0 T /2
e dt

A τ /2 − j nω0 t A 1 − j n ω0 t τ /2

T0 ∫−τ /2
= e dt = ⋅ ⋅ e
T0 − jnω0 −τ /2

A T0 ⎛ − j n ω0 τ2 τ
j n ω0 ⎞ A nπ τ
= ⋅ ⋅ ⎜e −e 2 ⎟ = ⋅ sin
T 0 − j 2π n ⎝ ⎠ nπ T0
1442443
τ 2π τ
= − j 2sin n ω0 = − j 2 sin n
2 T0 2

⎛ nπ τ ⎞
sin ⎜
Aτ ⎝ T0 ⎟⎠ Aτ nπ τ
= ⋅ = ⋅ sinc (A sinc function representation, equivalently)
T0 ⎛ nπ τ ⎞ T0 T0
⎜ T ⎟
⎝ 0 ⎠
Aτ +∞ ⎛ A nπ τ ⎞ j nω0 t Aτ +∞ ⎛ Aτ nπ τ ⎞ j nω0 t
f (t ) = +∑⎜ ⋅ sin ⎟ e or f (t ) = + ∑ ⎜ ⋅ sin c ⎟e
T0 n =−∞ ⎝ n π T0 ⎠ T0 n =−∞ ⎝ T0 T0 ⎠
n≠0 n≠0 8
(a). τ = 0.5 (b). τ = 0.25
T0 = 1 T0 = 1
A =1 A =1