Fourier Series Expansion
Fourier Series Expansion
T = 2/0 0 = fundamental frequency (rad/sec)
x t =
k =
ck e
jk t
0
Fourier Series (Complex Exponential Form)
These are 3 different forms for the same expression.
x t =a0 [ a k cos k0 t b k sin k0 t ]
k =1
Fourier Series (Trigonometric Form)
x t = A0 A k cos k0 t k
k =1
Fourier Series (Trigonometric Form)
x t =a0 [ a k cos k0 t b k sin k0 t ] =
k =1
k =
ck e
jk t
0
How to find ck and c-k from ak and bk?
1 jk t 1 jk t 1 jk t 1 jk t a k cos( k0 t )+b k sin( k0 t )=a k [ e + e ]+b k [ e e ] 2 2 2j 2j
0 0 0 0
a k b k jk t a k b k jk t =[ + ]e +[ ]e 2 2j 2 2j
0 0
ck 1 c k = (a k j b k ) 2 1 ck = (a k + j b k ) 2
ck
a k =c k +ck b k = j (c k ck )
k =1
x t = A0 A k cos k0 t k = c k e k =
jk t
0
Similarly by expanding out the kth term above using Euler's Formula you can see that
Ak j ck = e 2
Ak =2c k
k
Ak j ck = e 2
k =c k
Three (Equivalent) Forms of FS and Their Relationships
Trig Form: Sine-Cosine
a 0 =c 0 a k =c k ck , k =1,2,3, c 0=a0 b k = j c k ck , k =1,2,3,
x t =a 0 [ a k cos k 0 t b k sin k 0 t ]
k =1
ck = A0 =a 0 Ak = a 2 b 2 k k k =tan 1 a 0 = A0 a k =A k cos k b k =A k sin k
1 a jbk 2 k 1 a jbk 2 k c 0= A 0
k =1, 2, 3, k =1, 2, 3,
c k =
Exponential Form
x t =
k =
b k ak
cke
jk t
0
j 1 c k = Ak e k 2 j 1 c k = Ak e k 2
k =1, 2, 3, k =1, 2, 3, Ak =2c k A0=c 0 k =1, 2, 3, k =1, 2, 3,
Trig Form: Amplitude & Phase
k = c k
x t = A0 A k cos k 0 t k
k =1
Complex Exponential Form
x t =
k =
ck e
jk t
0
jk t 1 t T 0 c k = t x t e dt T
0 0
Fourier Series Trigonometric Form
f t =a0n=1 [a n cos n 0 t b n sin n 0 t ] 1 a 0= T f t dt T 2 a n= T f t cosn 0 t dt n0 T 2 b n= T f t sin n 0 t dt n0 T
We can derive these results in the same way as the complex exponential case using orthogonal functions!
f(t) = 1 for t = [-T/2, T/2]
Fourier Series Properties
Fourier Series Properties
if x (t )=a 0 +k =1 [ a k cos (k0 t )+b k sin (k0 t ) ] is real, then both a k and b k are real. And since ak j bk ck= 2 a k + j bk ck = 2
* k
then
c k =c
Hermitian Symmetry
i.e.,
c k=ck
c k =ck
Fourier Series
T = 2/0 0 = fundamental frequency (rad/sec)
x t =
k =
0 0
ck e
jk t
0
jk t 1 t T 0 c k = t x t e dt T
This is true for a very wide class of periodic signals! (see Dirichlet Conditions.)
Time shifting a periodic signal creates a new periodic signal! How are the Fourier series coefficients of the two signals related?
Fourier Series Properties
Time Shifting
f t c n jn t f tt 0 c n e
0 0
Time Differentiation
f t c n df t jn 0c n dt
Example
Find the Fourier series coefficients, dn, of h(t).
1 T /2 jn t d n= T /2 ht e dt T
0
1 T /4 T jn t d n= T /4 t e dt T 4
0
We have to integrate this by parts! Instead, we can use the differentiation theorem!
Example