Higher Engineering Mathematics, Sixth Edition

(Nancy Kaufman) #1

650 Higher Engineering Mathematics


thef(x) axis. Thus equation (15) could have been used,
giving:

cn=

2
L

∫ L
2
0

f(x)cos

(
2 πnx
L

)
dx

=

2
4

∫ 2

0

f(x)cos

(
2 πnx
4

)
dx

=

1
2

{∫ 1

0

5cos

(πnx
2

)
dx+

∫ 2

1

0dx

}

=

5
2




sin

(πnx
2

)

πn
2




1

0

=

5
2

(
2
πn

)(
sin


2

− 0

)

=

5
πn

sin


2

which is the same answer as in Problem 1; how-
ever, a knowledge of even functions has produced the
coefficient more quickly.

Problem 4. Obtain the Fourier series, in complex
form, for the square wave shown in Figure 71.4.

2

f(x)

0 x

22

2 ^2 ^3 

Figure 71.4

Method A
The square wave shown in Figure 71.4 is anodd
functionsince it is symmetrical about the origin.
The period of the waveform,L= 2 π.
Thus, using equation (16):

cn=−j

2
L

∫ L
2
0

f(x)sin

(
2 πnx
L

)
dx

=−j

2
2 π

∫π

0

2sin

(
2 πnx
2 π

)
dx

=−j

2
π

∫π

0

sinnxdx=−j

2
π

[
−cosnx
n


0

=−j

2
πn

(
(−cosπn)−(−cos0)

)

i.e. cn=−j

2
πn

[1−cosπn] (17)

Method B
If it hadnotbeen noted that the function was odd,
equation (12) would have been used, i.e.

cn=

1
L

∫ L
2
−L 2

f(x)e−j

2 πLnx
dx

=

1
2 π

∫π

−π

f(x)e−j

2 πnx
2 π dx

=

1
2 π

{∫ 0

−π

−2e−jnxdx+

∫π

0

2e−jnxdx

}

=

1
2 π

{[
−2e−jnx
−jn

] 0

−π

+

[
2e−jnx
−jn


0

}

=

1
2 π

(
2
jn

){[
e−jnx

] 0
−π


[
e−jnx


0

}

=

1
2 π

(
2
jn

){[
e^0 −e+jnπ

]

[
e−jnπ−e^0

]}

=

1
jπn

{
1 −ejnπ−e−jnπ+ 1

}

=

1
jnπ

{
2 − 2

(
ejnπ+e−jnπ
2

)}

by rearranging

=

2
jnπ

{
1 −

(
ejnπ+e−jnπ
2

)}

=

2
jnπ

{ 1 −cosnπ} from equation (3)

=

−j 2
−j(jnπ)

{ 1 −cosnπ}

by multiplying top and bottom by−j

i.e. cn=−j

2

(1−cosnπ) ( 17 )

It is clear that method A is by far the shorter of the two
methods.
Free download pdf