FOURIER SERIES

Philip Hall Jan 2011

Definition of a Fourier series
A Fourier series may be defined as an expansion of a function in a series
of sines and cosines such as

a0
f ( x) (an cos nx bn sin nx).
2 n 1

(7.1)

The coefficients are related to the periodic function f(x)

by definite integrals: Eq.(7.11) and (7.12) to be mentioned later on.
Henceforth we assume f satisfies the following (Dirichlet) conditions:
(1) f(x) is a periodic function;
(2) f(x) has only a finite number of finite discontinuities;
(3) f(x) has only a finite number of extrem values, maxima and minima in the
interval [0,2].
Fourier series are named in honour of Joseph Fourier (1768-1830), who made
important contributions to the study of trigonometric series, in connection with the
solution of the heat equation


a0
f ( x)
(a n cos nx bn sin nx).
2 n 1

(7.3)

Expressing cos nx and sin nx in exponential form, we may rewrite Eq.(7.1) as

1 inx inx
cos nx e e ,
2

f ( x)

inx
c
e
n

in which

and

1 inx inx
sinnx
e e
2i

1
cn (an ibn ),
2
1
c n (an ibn ),
2
1
c0 a0 .
2

(7.4)

(7.5)

n 0,
(7.6)

Completeness and orthogonality

We can show Fourier series satisfy
certain completeness properties- see next year

Now we need to find out how to determine the coefficients

in the Fourier series expansion and find out in what circumstances the
Fourier series and function coincide..

We can easily check the orthogonal relation for different values of the eigenvalue n by
choosing the interval 0,2

m ,n , m 0,
sin mx sin nxdx
m 0,
0,

m 0,
m,n ,
cos mx cos nxdx
2 , m n 0,

sin mx cos nxdx 0

(7.7)

(7.8)

for all integer m and n.

(7.9)

By use of these orthogonality, we are able to obtain the coefficients

a0
f ( x)
(a n cos nx bn sin nx).
2 n 1
multipling cos mx, and then integral from 0 to 2
2

a0
cos(
mx
)
f
(
x
)
dx

0
2

cos(mx)dx (a cos(nx) cos(mx)dx b sin(nx) cos(mx)dx)

0

n 1

(7.9)

Similarly
2

a0
sin(
mx
)
f
(
x
)
dx

0
2

sin(mx)dx (a cos(nx) sin(mx)dx b sin(nx) sin(mx)dx)

0

n 1

1 2
an f (t ) cos ntdt ,
(7.11)
0

1 2
bn f (t ) sin ntdt , n 0,1,2 (7.12)
0
Substituting them into Eq.(7.1), we write

(7.10)

1
f ( x)
2
1

2
2
1
f (t )dt (cos nx f (t ) cos ntdt sin nx f (t ) sin ntdt )
0
0
n 1
2

1
f (t )dt f (t ) cos n(t x)dt ,
n 1 0

(7.13)

This equation offers one approach to the development of the Fourier integral and
Fourier transforms.

EXAMPLE: Sawtooth wave

Let us consider a sawtooth wave
x,
0 x
x 2 , x 2 .

f ( x)

(7.14)

For convenience, we shall shift our interval from 0,2 to , . In this interval
we have simply f(x)=x. Using Eqs.(7.11) and (7.12), we have

an
bn

t cos ntdt 0,

t sin ntdt

t sin ntdt

t cos nt 0 cos ntdt

0
n
2
( 1) n1 ,
n
So, the expansion of f(x) reads

sin 2 x sin 3x
sin nx

f ( x) x 2 sin x

(1) n1
.
.
2
3
n

(7.15)

Figure 7.1 shows f(x) for the sum of 4, 6, and 10 terms of the series.
Three features deserve comment.
1.There is a steady increase in the accuracy of the representation as the number of
terms included is increased.
2.All the curves pass through the midpoint f ( x) 0 at x

Figure 7.1 Fourier representation of sawtooth wave

CONVERGENCE OF FOURIER SERIES

W assume f satisfies the following (Dirichlet) conditions:
(1) f(x) is a periodic function;
(2) f(x) has only a finite number of finite discontinuities;
(3) f(x) has only a finite number of extrem values,
maxima and minima in the
interval [0,2p].

We can then show that at a point where f(x) is

continuous the Fourier series converges to f(x). However
at a point of discontinuity the Fourier series converges to
(sum of right and left hand limits of limits of f(x). Hence at any point x

a0
f ( x) (an cos nx bn sin nx) .5( f f ).
2 n 1

(7.18)

where of course the right and left hand limits coincide where f is continuous.

7.2 ADVANTAGES, USES OF FOURIER SERIES

Discontinuous Function
One of the advantages of a Fourier representation over some other representation,
such as a Taylor series, is that it may represent a discontinuous function. An example
id the sawtooth wave in the preceding section. Other examples are considered in
Section 7.3 and in the exercises.
Periodic Functions
Related to this advantage is the usefulness of a Fourier series representing a periodic
,
functions . If f(x) has a period of 2 , perhaps it is only natural
that we expand it in
This guarantees that if
a series of functions with period 2 , 2 2 , 2 3
our periodic f(x) is represented over one interval 0,2 or , the
representation holds for all finite x.

At this point we may conveniently consider the properties of symmetry. Using the
interval , , sin x is odd and cos x is an even function of x. Hence ,
by Eqs. (7.11) and (7.12), if f(x) is odd, all an 0 if f(x) is even all bn 0 . In
other words,
a0
f ( x) an cos nx, f (x) even, (7.21)
2 n1

f ( x) bn sin nx,

f (x)

odd.

(7.21)

n 1

Frequently these properties are helpful in expanding a given function.

We have noted that the Fourier series periodic. This is important in considering
whether Eq. (7.1) holds outside the initial interval. Suppose we are given only that

f ( x ) x,

0 x

(7.23)

and are asked to represent f(x) by a series expansion. Let us take three of the
infinite number of possible expansions.

1.If we assume a Taylor expansion, we have

f ( x ) x,

(7.24)

a one-term series. This (one-term) series is defined for all finite x.

2.Using the Fourier cosine series (Eq. (7.21)) we predict that

f ( x ) x,
f ( x) 2 x,

x 0,

x 2 .

(7.25)

3.Finally, from the Fourier sine series (Eq. (7.22)), we have

f ( x ) x,
f ( x) x 2 ,

x 0,
x 2 .

(7.26)

Figure 7.2 Comparison of Fourier cosine series, Fourier sine series and Taylor series.

These three possibilities, Taylor series, Fouries cosine series, and Fourier sine series,
are each perfectly valid in the original interval 0, . Outside, however, their behavior
is strikingly different (compare Fig. 7.3). Which of the three, then, is correct? This
question has no answer, unless we are given more information about f(x). It may be
any of the three ot none of them. Our Fourier expansions are valid over the basic
interval. Unless the function f(x) is known to be periodic with a period equal to our
basic interval, or (1 n) th of our basic interval, there is no assurance whatever that
representation (Eq. (7.1)) will have any meaning outside the basic interval.
It should be noted that the set of functions cos nx , n 0,1,2 , forms a
complete orthogonal set over 0, . Similarly, the set of functions sin nx , n 1,2,3
forms a complete orthogonal set over the same interval. Unless forced by boundary
conditions or a symmetry restriction, the choice of which set to use is arbitrary.

FOURIER SERIES OVER ARBITRARY INTERVALS

So far attention has been restricted to an interval of length of 2 . This restriction
may easily be relaxed. If f(x) is periodic with a period 2 L , we may write

a0
nx
nx
f ( x) an cos
bn sin
,

2 n1
L
L

(7.27)

with

1 L
nt
an f (t ) cos
dt ,

L
L
L
1 L
nt
bn f (t ) sin
dt ,

L
L
L

(7.28)

n 0,1,2,3 ,
n 0,1,2,3 ,

(7.29)

replacing x in Eq. (7.1) with x L and t in Eq. (7.11) and (7.12) with t L
(For convenience the interval in Eqs. (7.11) and (7.12) is shifted to t . )
The choice of the symmetric interval (-L, L) is not essential. For f(x) periodic with
a period of 2L, any interval ( x0 , x0 2 L) will do. The choice is a matter of
convenience or literally personal preference.

HALF RANGE FOURIER SERIES

We have seen above that even and odd functions have Fourier series
With either an . or
bn .zero.
Now suppose we have a function defined on an interval [0,K], then
we extend it as an even or odd function of period K so as to produce a Fourier
cosine or sine series. But note the period is now 2K, it is always wise
to sketch the extended function.
Exercise: extend the functions given below as odd and even functions,
sketch the extended functions and give the formulae for the Fourier coefficients:

(a) f ( x) sin x,0 x , (b) f ( x) x,0 x 1.

(7.4-5)

Note that extending a function to produce either an odd or even function

can lead to discontinuities.

7.3 APPLICATION OF FOURIER SERIES

Example 7.3.1 Square Wave High Frequency
One simple application of Fourier series, the analysis of a square wave (Fig. (7.5))
in terms of its Fourier components, may occur in electronic circuits designed to
handle sharply rising pulses. Suppose that our wave is designed by

f ( x) 0,
f ( x) h,

x 0,
0 x .

(7.30)

From Eqs. (7.11) and (7.12) we find

1
a0 hdt h,
0
1
an h cos ntdt 0
0
1
bn

(7.31)
(7.32)

2h
h

, n odd
h sin ntdt
(1 cos n ) n
;
n
0, n even.

(7.33)

The resulting series is

f ( x)

h
2

2h sin x sin 3 x sin 5 x

(

).

1
3
5

(7.36)

Except for the first term which represents an average of f(x) over the interval ,
all the cosine terms have vanished. Since f ( x) h 2 is odd, we have a Fourier sine
1
series. Although only the odd terms in the sine series occur, they fall only as n
This is similar to the convergence (or lack of convergence ) of harmonic series.
Physically this means that our square wave contains a lot of high-frequency
components. If the electronic apparatus will not pass these components, our square
wave input will emerge more or less rounded off, perhaps as an amorphous blob.

Example 7.3.2 Full Wave Rectifier

As a second example, let us ask how the output of a full wave rectifier approaches
pure direct current (Fig. 7.6). Our rectifier may be thought of as having passed the
positive peaks of an incoming sine and inverting the negative peaks. This yields

f ( x) sin t ,
f ( x) sin t ,

0 t ,
t 0.

(7.37)

Since f(t) defined here is even, no terms of the form sin nt will appear.
Again, from Eqs. (7.11) and (7.12), we have

1 0
1
a0 sin td (t )

2
4
sin td (t ) ,
0

sin td (t )
(7.38)

2
an sin t cos ntd (t )
0
2 2

,
n even
2
n 1
0
n odd .

(7.39)

Note carefully that 0, is not an orthogonality interval for both sines and cosines
together and we do not get zero for even n. The resulting series is

2 4 cos nt
f (t )
.

2
n2, 4,6, n 1

(7.40)

The original frequency has been eliminated. The lowest frequency oscillation is
2 The high-frequency components fall off as n 2 , showing that the full wave
rectifier does a fairly good job of approximating direct current. Whether this good
approximation is adequate depends on the particular application. If the remaining ac
components are objectionable, they may be further suppressed by appropriate filter
circuits.

These two examples bring out two features characteristic of Fourier expansion.
1. If f(x) has discontinuities (as in the square wave in Example 7.3.1), we can expect
the nth coefficient to be decreasing as 1 n . Convergence is relatively slow.
2. If f(x) is continuous (although possibly with discontinuous derivatives as in the
Full wave rectifier2of example 7.3.2), we can expect the nth coefficient to be
decreasing as 1 n
3. More generally if f and its first r derivatives are continuous, but the r+1 is not
2
then the nth coefficient will be decreasing as 1 n

Example 7.3.3 Infinite Series, Riemann Zeta Function

As a final example, we consider the purely mathematical problem of
expanding f ( x ) x 2 ,
(7.41)
x

by symmetry all bn 0 For the

1
a0

an s we have

2
2

2
x
dx 3 ,

2 2
an x cos nxdx
0
2
2
(1) n 2

n
4
(1) n 2 .
n

(7.42)

(7.43)

From this we obtain

2

2
n cos nx
x
4 ( 1)
.
2
3
n
n 1

(7.44)

As it stands, Eq. (7.44) is of no particular importance, but if we set

cos n (1) n

(7.45)

and Eq. (7.44) becomes

2
1

4 2
3
n 1 n
2

(7.46)

or

2 1
2 (2),
6 n1 n

(7.47)

thus yielding the Riemann zeta function, (2) , in closed form. From our
2
expansion of x and expansions of other powers of x numerous other
infinite series can be evaluated.

Some examples of Fourier series

1.
2.

12 ( x), x 0
1
sin nx 1

0 x
n 1 n
2 ( x),

(1) n 1
n 1

1
sin nx
n

1
2

x,

3.

4, x 0
1
sin(
2
n

1
)
x

0 x
2
n

1
n 0
4,

4.

x
1
cos nx ln 2 sin( ) , x

2
n 1 n

5.

n 1
(

1
)
cos
nx

ln
2
cos(
) , x

n
2

n 1

6.

x
1
1
cos(2n 1) x ln cot( ) , x

2
2
n 0 2n 1

7.4 Properties of Fourier Series

Convergence
It might be noted, first that our Fourier series should not be expected to be
uniformly convergent if it represents a discontinuous function. A uniformly
convergent series of continuous function (sinnx, cosnx) always yields a continuous
function. If, however,
(a) f(x) is continuous, x
(b)

f ( ) f ( )

(c)

f (x) is sectionally continuous,

the Fourier series for f(x)

) will converge uniformly. These restrictions do not demand
that f(x) be periodic, but they will satisfied by continuous, differentiable, periodic
function (period of 2

Integration
Term-by-term integration of the series

a0
f ( x)
(a n cos nx bn sin nx).
2 n 1

(7.60)

yields

x0

f ( x)dx

a0 x
a
b
n sin nx n cos nx .
2 x0 n 1 n
n 1 n
x0
x0

(7.61)

Clearly, the effect of integration is to place an additional power of n in the

denominator of each coefficient. This results in more rapid convergence than
before. Consequently, a convergent Fourier series may always be integrated
term by term, the resulting series converging uniformly to the integral of the
original function. Indeed, term-by-term integration may be valid even if the
original series (Eq. (7.60)) is not itself convergent! The function f(x) need only
be integrable.

Strictly speaking, Eq. (7.61) may be a Fourier series; that is , if a0 0

there will be a term

x0

1
2

a0 x . However,

f ( x ) 12 a0 x

(7.62)

will still be a Fourier series.

Differentiation
The situation regarding differentiation is quite different from that of integration.
Here thee word is caution. Consider the series for

f ( x ) x,

(7.63)

We readily find that the Fourier series is

x 2 (1) n 1
n 1

sin nx
, x
n

(7.64)

Differentiating term by term, we obtain

1 2 (1) n 1 cos nx,

(7.65)

n 1

which is not convergent ! Warning. Check your derivative

For a triangular wave which the convergence is more rapid (and uniform)

4 cos nx
f ( x)
.

2
2 n 1, odd n

(7.66)

Differentiating term by term

4 sin nx
f ( x)
.

n 1, odd n

(7.67)

which is the Fourier expansion of a square wave

0 x ,
1,
1, x 0.

f ( x)

(7.68)

As the inverse of integration, the operation of differentiation has placed an

additional factor n in the numerator of each term. This reduces the rate of
convergence and may, as in the first case mentioned, render the differentiated
series divergent.
In general, term-by-term differentiation is permissible under the same conditions
listed for uniform convergence.

COMPLEX FORM OF FOURIER SERIES

Recall from earlier that we can write a Fourier series in the complex form

f ( x)

inx
c
e
n

in which

and

1
cn (an ibn ),
2
1
c n (an ibn ),
2
1
c0 a0 .
2

From the earlier definitions of

1
cn
T

n 0,

an , bn

we can show that

f ( x)e in x dx, 2 / T ,

The multiplication theorem and Parsevals theorem

From the complex form of the definition of Fourier series for functions
f(x), g(x) of period 2T we can show that

cn d

*
n

f ( x) g ( x)dx,,

Where the c,ds are the Fourier coefficients of f,g respectively.

If we use the repalce the cds by their real form and take f=g
we deduce Paresevals theorem
,
,

1 2
2 T
2
2
a 0 a n b n f ( x ) f * ( x )dx,,
2
T 0
1