Abstract

This project studies mainly the design techniques of wideband circularly polarized patch

antenna with broadened beamwidth. Through combining different features, an axial ratio

bandwidth (<3dB) of 3.2% can be achieved within the impedance bandwidth with a broad

beamwidth of around 120 degrees. The dominant radiating mode consists of Left Hand

Circularly Polarized (LHCP) fields.

The main objectives of designing a patch antenna include broadening the impedance

bandwidth – which is an inherent disadvantage of patch antenna – and moreover, the

axial ratio bandwidth if the antenna is to work circularly polarized. To achieve this, an

imbalanced U-slot is adopted together with truncated corners on the radiating plate.

Another goal is to broaden the beamwidth, realized by forming metallic walls around the

main patch.

Following the introduction, systematic comparisons are made in this report regarding

each important parameter of the antenna by varying a designated dimension while keep

others unchanged. Then, the resultant antenna will be presented and lastly, some

techniques for further improvement will be suggested.

All the simulations in this report are done using Zeland IE3D.

1
 Section 1

INTRODUCTION

2
1.1 Background of the Project

Antenna is specifically designed to send out and receive electromagnetic signals in space

and thus can convert electromagnetic waves into electrical signals such as current and the

other way around. It is used widely in telecommunications.

There are many classes of antenna, one of which is the patch antenna. Patch antenna has

lots of advantages such as low cost, easy to fabricate, light weight and polarization

diversity. On the other hand however, it has also got some inherent disadvantages such as

poor matching performance, low impedance bandwidth and axial ratio bandwidth if the

antenna is circularly polarized.

Therefore, the major goal of designing an antenna is to develop techniques to compensate

for those disadvantages. This project adopted two measures and the details will be dealt

with in later sections.

The beamwidth is the angle between -3dB points of the main lobe in an antenna’s

radiation pattern; clearly it describes the effective coverage of the antenna in space as

within this range, the radiated power does not decrease rapidly. In some application,

larger beamwidth is preferred because the in certain communication, large coverage of

geographical area is required. One thing to note is that if the input powers to antenna are

the same, an antenna with wider beamwidth will have weaker radiation within the

beamwidth compared to the radiated power of narrow beamwidth antenna in average.

3
1.2 Introduction to Antenna Fundamentals

Lots of parameters are of concern when evaluating the antenna performance and they will

be discussed in order below.

1.2.1 Impedance matching

For a patch antenna, it is usually fed by a coaxial cable. And normal characteristic

impedance of a coaxial cable is 50Ω. From the theory of transmission line, all the power

will be transferred from the input to the load ideally if the impedance of the load matches

the characteristic impedance of the transmission cable, which is 50Ω.If the input

impedance of the load deviates from the standard 50Ω, then some power will be reflected

back in the form of a wave. Antenna can just be considered as such kind of load, except

that its input impedance will vary with respect to the frequency of the input signal.

1.2.2 Return loss S(1,1)

The return loss is just the power of the reflected wave over total input power under

mismatch conditions. Usually dB is used to denote the return loss and conventional

threshold for return loss is -10dB, below which more than 90% of the input power is

transmitted to the antenna and thus less than 10% of the total power is reflected back.

1.2.3 Polarization

Three kinds of polarization exist when the wave propagates in space: linearly, circularly

and elliptically polarized. In this project, circularly polarized antenna is to be designed

and circular polarization means that the tip of the E-field vector forms a circle as time

elapses, at a predefined point in space.

4
Moreover, circular polarization can be further divided into two categories: left-hand CP

and right-hand CP, determined by the direction the circle forms (clockwise or

counterclockwise) with respect to the propagation direction.

1.2.4 Axial ratio

It is a coefficient to describe the polarization purity in communication. Consider an

antenna previously designed to send out left-hand CP wave, however, due to various

factors such as surface perturbations, right-hand CP wave will also be generated. Of

course it is preferable to control the signal strength of right-hand CP wave under a certain

level compared with left-hand CP wave, therefore axial ratio is produced.

The axial ratio can be calculated using the power of left-hand CP wave and right-hand CP

wave in dB scale, using equation (1). Usually an axial ratio bandwidth is defined as the

frequency band in which the axial ratio is below 3dB.

P L + PR
AR ( dB )=20 log | |
PL −P R
… …(1)

1.2.5 Bandwidth

Bandwidth is one of the most important features to evaluate the performance of an

antenna. From information and communication theories, it is known that a larger

bandwidth brings higher data transmission rate, which is desirable for modern

communication systems. Therefore, enlarging bandwidth of antenna is always a

fundamental requirement of designers.

5
The bandwidth is defined as the frequency range between the lower and upper frequency

specified by different requirements.

1.2.5.1 Impedance bandwidth

The impedance bandwidth is defined as the frequency band whose return loss is smaller

or equal than -10dB, or S(1, 1)≤-10dB as shown in the figure below, fl stands for the

lower frequency and fu for upper frequency. Clearly this takes the efficiency of antenna

into the account.

Figure 1.1 - Return loss versus frequency

For patch antenna, the bandwidth is usually represented as percentage bandwidth, given

by equation (2).

2 ×(f u −f l )
B .W .= × 100 % … …(2)
f u+ f l

6
1.2.5.2 Axial ratio bandwidth (ARBW)

Similar to the definition of impedance bandwidth, for circularly polarized antenna, axial

ratio bandwidth is defined as the frequency band whose axial ratio is smaller or equal

than 3dB as shown in the figure 1.2.

Figure 1.2 – Axial ratio versus frequency

With the lower frequency being fl and upper frequency being fu, the axial ratio bandwidth

can be calculated similarly using equation (2).

1.2.6 Radiation pattern

Radiation pattern means how the radiation field (usually power) changes with the

direction, and is commonly drawn in polar coordinate. For radiation pattern, it is

favorable that the pattern will not change significantly with frequency within the

bandwidth, so that the radiation will be stable when frequency alters in its bandwidth.

7
For each radiation pattern, there are many lobes and the main lobe is the lobe that

contains the most radiation power and are of most concern while other lobes are side

lobes or back lobes.

1.2.7 Beamwidth

Considering the main lobe in an antenna’s radiation pattern, the beamwidth is defined as

the angle between the -3dB points. Taking the following figure[1] as an example.

Figure 1.3 – Example illustrating beamwidth

The beamwidth is the angle between the two orange straight lines and from the figure, the

beamwidth is approximately 50°.

8
 Section 2

STUDY OF ANTENNA

PARAMETERS

9
In the last section, the indicators of evaluating the performance of the antenna are

introduced and in this section; and here parameters of the patch antenna will be discussed

in details to see how they will influence the performance of the antenna.

The study will focus mainly on antenna patch size, size of the ground plane, feeding point

position, length of truncated corners, dimensions of the U-slot and specifications of

circular walls around the patch. Comparisons will be made for each parameter in terms of

impedance bandwidth, axial ratio bandwidth and beamwidth.

The results from this section are all obtained by simulation using Zeland IE3D software.

2.1 Antenna patch size

In this project, the substrate is air and therefore the dielectric constant ξ=1. With a given

central frequency, the patch size of the antenna can be approximately calculated using the

following equation:

c
L≈ … …(3)
2 f √ εr

where f is the theoretical resonant frequency, c is the speed of light, L is the length of the

patch and ε r is the relative permittivity.

For the square patch antenna of this project, the resonance frequency is set to around

f0=2.3GHz and from the equation, L is 65mm. The actual length of the patch however, is

around 45mm considering the slotting of the patch and the fact that the equation only

gives an approximation of the dimension of the radiating plate.

10
 L
L
L

(0,y)

Figure 2.1 – Illustration of the studied antenna

To study the effect of varying patch size to the performance of the antenna, the length L

shown in figure 2.1 will be increased from 42.7mm to 46.7mm in a step size of 1mm.

The results are shown in the following table:

Table 2.1 – Performance of antenna with different patch sizes

L (mm) Impedance bandwidth Axial ratio bandwidth Beamwidth

(GHz) (GHz) (degrees)

42.7 2.366-2.282=3.61% 0 115
43.7 2.391-2.265=5.41% 2.380-2.356=1.01% 119
44.7 2.361-2.251=4.77% 2.342-2.275=2.91% 120
45.7 2.334-2.246=3.84% 2.265-2.229=1.6% 122
46.7 2.301-2.234=2.95% 0 116

11
 From the above table, it can be seen that the working frequency band shifts to the left

when the patch size increases. This agrees with equation (3), also both the axial ratio

bandwidth and impedance bandwidth decrease when L goes away from around 45mm.

The workable frequency range of a circular polarized antenna should be the overlapped

part of impedance bandwidth and axial ratio bandwidth, thus 44.7mm is clear to be the

optimal patch size as it gives the largest working bandwidth. On the other hand, because

the patch size has an impact on how close the radiating plate is to the surrounding walls,

the beamwidth also changes.

2.2 Feeding point position

For the antenna shown in figure 2.1, if the center of the ground plane is set as (0,0), the

position of the feeding probe (0, y) is to be studied. y in this case will vary from 3.45 to

11.45 with a step size of 2.

Table 2.2 – Performance of antenna with different feeding positions

Feeding point Impedance bandwidth Axial ratio bandwidth Beamwidth

position (0,y) (GHz) (GHz) (degrees)

(0, 3.45) 2.343-2.258=3.69% 0 120
(0, 5.45) 2.353-2.255=4.25% 2.338-2.287=2.2% 120
(0, 7.45) 2.361-2.251=4.77% 2.342-2.275=2.91% 120
(0, 9.45) 2.377-2.248=5.58% 0 120
(0, 11.45) 2.39-2.25=6.03% 0 120

From the results, it can be concluded that the impedance bandwidth actually increases

when the position of the feeding point moves upwards, the axial ratio bandwidth

however, first increases and then goes back to zero after reaching its maximum at y=7.45.

The beamwidth at this stage does not change a lot.

12
2.3 Length of truncated corners

The length of truncated corners is represented as d shown in the following figure.

a
b

d c

Figure 2.2 – Illustration of the studied antenna (Zoomed)

The length m will increase from 5.6mm to 7.6mm in a step size of 0.5mm to see how it

affects the performance of the antenna.

Table 2.3 – Performance of antenna with different lengths of truncated corners

m (mm) Impedance bandwidth Axial ratio bandwidth Beamwidth

(GHz) (GHz) (degrees)

5.6 2.351-2.23=5.28% 2.34-2.284=2.42% 120
6.1 2.358-2.24=5.13% 2.341-2.281=2.6% 120
6.6 2.361-2.251=4.77% 2.342-2.275=2.91% 120
7.1 2.372-2.266=4.57% 2.344-2.277=2.9% 120

13
 7.6 2.376-2.276=4.3% 2.346-2.278=2.94% 119

From the table, we can see that the impedance bandwidth and axial ratio bandwidth

actually follows an opposite trend of change when the length increases.

2.4 Dimensions of the U-slot

Referring to figure 2.2, four parameters a, b, c and d will be dealt with separately in the

following part.

2.4.1 Length of the left slot – a

The length a will be spread from 25.7mm to 29.7mm in a step size of 1mm.

Table 2.4 – Performance of antenna with different values of a

a (mm) Impedance bandwidth Axial ratio bandwidth Beamwidth

(GHz) (GHz) (degrees)

25.7 2.386-2.308=3.32% 0 120
26.7 2.373-2.283=3.87% 2.328-2.296=1.38% 120
27.7 2.361-2.251=4.77% 2.342-2.275=2.91% 120
28.7 2.356-2.218=6.03% 0 120
29.7 2.349-2.185=7.23% 0 120

From the results, it can be concluded that the impedance bandwidth actually increases

when the length of the left leg of the U-slot gets larger, the axial ratio bandwidth

14
however, reaches its maximal point at a=27.7mm. The beamwidth does not change a lot

at this stage.

2.4.2 Length of the right slot – b

The length a will be spread from 14.3mm to 18.3mm in a step size of 1mm.

Table 2.5 – Performance of antenna with different values of b

b (mm) Impedance bandwidth Axial ratio bandwidth Beamwidth

(GHz) (GHz) (degrees)

14.3 2.386-2.28=4.54% 2.344-2.294=1.08% 120
15.3 2.378-2.26=5.09% 2.344-2.291=2.29% 120
16.3 2.361-2.251=4.77% 2.342-2.275=2.91% 120
17.3 2.348-2.246=4.44% 2.342-2.275=2.91% 120
18.3 2.326-2.246=3.5% 2.338-2.272=2.86% 120

From the above results, it can be discovered that the impedance bandwidth and axial ratio

bandwidth are not so sensitive to the change of b, especially when b is in the range of

15.3mm to 18.3mm although the impedance bandwidth decreases slightly. Similarly the

beamwidth does not change with respect to b.

2.4.3 Width of the U slot – c

The width of the U-slot will vary from 16.4mm to 20.4mm with a step size of 1mm and

the results are shown in the following table.

Table 2.6 – Performance of antenna with different values of c

c (mm) Impedance bandwidth Axial ratio bandwidth Beamwidth

(GHz) (GHz) (degrees)

16.4 2.427-2.342=3.56% 2.335-2.28=2.38% 120
17.4 2.396-2.294=4.35% 2.338-2.275=2.73% 120
18.4 2.361-2.251=4.77% 2.342-2.275=2.91% 120

15
 19.4 2.32-2.227=4.09% 2.347-2.285=2.68% 120
20.4 2.256-2.215=1.83% 2.348-2.288=2.59% 120

The results show that the axial ratio actually stays nearly the same with different widths

of the U-slot, the impedance bandwidth however, shifts to lower frequency band when c

increases. What’s more, for some values of c (16.4mm, 20.4mm), the axial ratio

bandwidth totally does not overlap with impedance bandwidth, resulting in no net

working bandwidth.

2.4.4 Position of the U slot – d

The distance of the U-slot from the edge of the patch d, referring again to figure 2.2, is

varied from 6.3mm to 10.3mm with a step size of 1mm.

Table 2.7 – Performance of antenna with different values of d

d (mm) Impedance bandwidth Axial ratio bandwidth Beamwidth

(GHz) (GHz) (degrees)

6.3 2.374-2.298=3.25% 2.304-2.297=0.3% 120
7.3 2.371-2.258=4.88% 2.332-2.282=2.17% 120
8.3 2.361-2.251=4.77% 2.342-2.275=2.91% 120
9.3 2.363-2.254=4.72% 2.342-2.275=2.91% 120
10.3 2.365-2.258=4.63% 2.343-2.281=2.68% 120

It can be concluded that both bandwidths do not change significant with different values

of d within a certain range, out of which the working frequency band drops quickly.

2.4 Size of the ground plane

16
As the ground plane is circular as shown in figure 2.1, different radius R of the plate is

studied. The aim of the project is to reach a beamwidth of around 120 degrees and three

radius (R=90mm, 105mm and 120mm) are studied in detail.

Table 2.8 – Performance of antenna with different values of R

R (mm) Impedance bandwidth Axial ratio bandwidth Beamwidth

(GHz) (GHz) (degrees)

100 2.356-2.248=4.7% 2.342-2.274=2.95% 116
110 2.36-2.251=4.73% 2.342-2.275=2.95% 117
120 2.361-2.251=4.77% 2.342-2.275=2.91% 120

As from the results, it can be concluded that the size of the ground plane actually does not

influence a lot on impedance and axial ratio bandwidth. But it does affect beamwidth of

the antenna as it acts as a component to shape the radiated fields.

2.5 Radius of circular walls

Let r be the radius of the most inner wall, r will be given three values: 34mm, 39mm and

44mm to see the effect of circular walls around the main patch.

Table 2.9 – Performance of antenna with different values of r

r (mm) Impedance bandwidth Axial ratio bandwidth Beamwidth

(GHz) (GHz) (degrees)

34 2.361-2.251=4.77% 2.342-2.275=2.91% 120
39 2.345-2.25=4.13% 2.4-2.337=2.66% 109
44 2.345-2.254=3.96% 2.4-2.335=2.75% 99

Clearly the radius of the circular walls influences a lot over the beamwidth of the

antenna, the rationale lies in the fact that the walls produce scattered field to ensure that

17
the field strength does not decrease so quickly in the expected range. Moreover, the

results above state that smaller the radius, larger the beamwidth. This is easy to

understand because as the radius becomes large, the effect is weaker in producing

scattered field.

18
 Section 3

DESIGN AND

SIMULATION RESULTS

19
As there are mainly three stages for my design, this section will be divided into three

parts, each introducing the specifications of the antenna together with the performance

from simulation.

3.1 Design stage one – Patch antenna with unbalanced U-slot

At this stage, the main goal is to achieve circular polarization of the antenna. The reason

why circular polarization is preferred over linear polarization is because it can avoid the

problem of heavy absorption and multipath reflection. In telecommunication, the cloud,

rain or snow may act as an anisotropic material thus absorb the electromagnetic wave in a

certain polarization more than others. Circularly polarization however, rotates in

polarizations and therefore will not be severely affected when this situation occurs. The

problems related to multipath reflection actually have similar rationales considering the

fact that transmission media will possibly reflect the wave in a given direction more than

others.

Circular polarization is usually achieved by two orthogonal plane waves with equal

magnitude and 90 degrees phase difference. A lot of methods have been developed up till

now to meet this requirement. For example, in the early days, two orthogonal radiating

strips were adopted and dual feed structure was utilized to input 90 degrees out of phase

current. Its performance however, is not so satisfactory because it is really hard to input

current exactly with phase difference of 90 degrees and the lengths of the radiating strips

should also be taken into consideration.

20
In this design, an unbalanced U-slot will be cut in the patch of the antenna. Unbalanced

here means that two branches of the slot are not equal in length, this will cause a phase

different between the orthogonal radiated wave components, leading to circular

polarization. On the other hand, U-slot can sometimes act to increase the impedance

bandwidth because it introduces multi-resonances so that when two resonant frequencies

are close to each other, the impedance bands brought about by them will be continuous

and hence impedance bandwidth is broadened.

In the beginning of the project, the relative permittivity of dielectric substrate is ε r=2.2.

The structures of the antenna are shown in figures below, followed by a table describing

the specifications of the geometry.

L

n
a
b

d c

Figure 3.1 – 2D view of the antenna and the labeling of parameters of design 1

21
 Figure 3.2 – 3D view of the antenna of design 1

The dimensions which are labeled were tuned so that the antenna reaches optimal

performance, that is, high impedance bandwidth as well as maximal axial ratio

bandwidth. The results obtained are listed in table 1 as follows.

Table 3.1 – Parameters of unbalanced U-slot antenna

Parameters Explanation Length (mm)

L Ground plane size 100
w Radiation patch size 44.7
h Height of the radiating plate 11
a Length of longer edge of U-slot 28
b Length of shorter edge of U-slot 22
c Distance of U-slot from bottom 8.3
d Width of U-slot 17
g Width of the slot cut 2.3
n Distance of feed from the origin 8.45

The simulation results of the first design are shown as follows. Change of returns loss

S(1,1) versus the frequency is displayed first in order to calculate the impedance

bandwidth.

22
 Figure 3.3 Return loss versus frequency for design 1

The bandwidth therefore can be calculated easily:

2 ×(1.7408−1.6208)
B .W .= ×100 %=7 %
1.7408+1.6208

The following figure deals with axial ratio bandwidth.

23
 Figure 3.4 Axial ratio versus frequency for design 1

2 ×(1.6704−1.6288)
ARBW = × 100 %=2.5 %
1.6704+1.6288

Lastly for beamwidth, as there is no measurement to enlarge it, the beamwidth will be in

its normal range, below 90 degrees. Nevertheless, verification is still preferred and

therefore the radiation pattern of the antenna is shown as follows, but with only left hand

polarized component which is the main radiated field.

24
 Figure 3.5 Radiation pattern for design 1

Beamwidth=44.24 ° +33.66 ° =80.9°

From the calculation, the beamwidth only reaches 80.9˚, which is just as expected. For

applications requiring large geographical coverage, this beamwidth is far from enough.

25
3.2 Design stage two –Unbalanced U-slot with truncated corners

This stage aims mainly on broadening the bandwidth, especially the axial ratio bandwidth

of the antenna. As mentioned in the last part, lots of methods can be used to achieve

circular polarization. And to truncate corners in the radiating patch is one of them. A

reasonable guess would be to add these two features together to strengthen the effect of

circularly polarization thus widen the axial ratio bandwidth.

The permittivity of the dielectric substrate is still ε r=2.2 for this stage and the following

is the structure of the antenna with the listing of parameters.

L

n
b

m d c

Figure 3.6 2D view of the antenna and the labeling of parameters of deign 2

The height of the patch is still h identical to the last part and is not repeatedly shown here.

26
 Figure 3.7 – 3D view of the patch antenna of design 2

Table 3.2 – Parameters of unbalanced U-slot antenna with truncated corners

Parameters Explanation Length (mm)

L Ground plane size 100
w Radiation patch size 44.7
h Height of the radiating plate 11
a Length of longer edge of U-slot 29.3
b Length of shorter edge of U-slot 17.8
c Distance of U-slot from bottom 8.3
d Width of U-slot 18
g Width of the slot cut 2.3
m Length of truncated corner 6.6
n Distance of feed from the origin 7.45

The simulated return loss S(1,1) with respect to the frequency is shown below in figure

3.8. From the figure, the impedance bandwidth can also be calculated.

2 ×(1.762−1.6276)
B .W .= ×100 %=8 %
1.762+1.6276

which is better compared with the original 7.16%.

27
 Figure 3.8 – Return loss versus frequency for design 2

The axial ratio bandwidth can be calculated from figure 3.9, describing the axial ratio in

various frequencies.

2 ×(1.7046−1.6514)
ARBW = × 100 %=3.2 %
1.7046+1.6514

which is a large improvement compared with 2.5% in the previous section. The

percentage increase is around 28%. This justifies the guess stated at the beginning of this

part, by adding different features of circular polarization together properly, the axial ratio

bandwidth can be improved.

28
 Figure 3.9 – Axial ratio versus frequency for design 2

Because in this part, still no measurement is taken to broaden the beamwidth of the

antenna, the -3dB points in the radiation pattern correspond to -38.53˚ and 41.56˚

respectively. This gives a beamwidth of 80.09˚ which is nearly the same as that of the

previous antenna without truncated corners.

3.3 Design stage three – Circular walls surrounding radiation patch

The final stage is to achieve the feature of large beamwidth (120˚) and this can be

realized using surrounding walls around the main radiation patch. In the previous part,

solid substrate is used, in the final stage however, to facilitate the process of placing

29
metal walls around the patch and lower the cost of the antenna, air substrate is utilized. A

possible disadvantage of using air substrate will be a very large ground plane size.

The structure of the antenna is shown below in two separate figures for easier labeling of

parameters.

R n
a
r b
e

m d c

(a) (b)

Figure 3.10 – 2D view of the antenna and the labeling of parameters of design 3

(a) whole antenna (b) zoomed version for main radiation patch

30
 H

Figure 3.11 – 3D view of the patch antenna of design 3

Table 3.3 – Parameters of antenna with surrounding circular walls

Parameters Explanation Length (mm)

R Ground plane radius 120
r Surrounding wall radius 34
f Distance between walls 5
w Radiation patch size 44.7
H Height of the circular walls 18
h Height of the radiating plate 10
a Length of longer edge of U-slot 27.7
b Length of shorter edge of U-slot 16.3
c Distance of U-slot from bottom 8.3
d Width of U-slot 18.4
e Width of the bottom slot cut 0.9
g Width of the side slot cut 2.3
m Length of truncated corner 6.6
n Distance of feed from the origin 7.45

The impedance bandwidth can be calculated according to figure 3.12, showing the return

loss of the antenna.

2 ×(2.35592−2.2472)
B .W .= ×100 %=4.723%
2.35592+2.2472

31
Clearly the walls have a large impact over the impedance matching of the antenna and the

impedance bandwidth reduced a lot. But it is not problematic if the axial ratio bandwidth

does not change a lot as the working frequency band is the overlapped range between the

two.

Figure 3.12 – Return loss versus frequency for design 3

The axial ratio bandwidth can be calculated similarly referring to figure 3.13.

2 ×(2.35592−2.2868)
ARBW = ×100 %=2.9 %
2.35592+2.2868

which is not a very large reduction and lies in the range of impedance bandwidth. Such

axial ratio bandwidth is enough for applications requiring circular polarization.

32
 Figure 3.13 – Axial ratio versus frequency for design 3

Lastly for the beamwidth, it can be calculated from the radiation pattern shown below.

Beamwidth=66.0432 °+54.9435 °=120.9867 °

It is a clear improvement compared with beamwidths in the previous parts, which are

around 80˚. The results prove the validity of enlarging the beamwidth by adding circular

walls around the radiating patch.

33
Figure 3.14 – Radiation pattern for design 3

34
 Section 4

ANTENNA FABRICATION

AND MEASUREMENT

Only the finalized design, in which the patch is surrounded with walls, was fabricated.

The antenna is fabricated without using addition substrate with air as the only substrate.

35
This makes the cost of fabrication rather cheap, which is certainly an advantage. The

following figure shows the real product of the antenna.

Figure 4.1 – Real fabricated product

The ground plane of the antenna used aluminum as the material whereas the patch and

walls of the antenna adopted copper, which is thinner compared with aluminum. The

radius of the feeding probe to the patch is 0.5mm and there is a connector at the back side

of the ground plate.

For antenna measurement, it is mainly divided into two parts for my project.

The first part deals with the impedance bandwidth of the antenna. This can be achieved

by using Network Analyzer. Network Analyzer will give the reading of return loss S(1,1)

36
versus frequency in a given range very quickly. With the data obtained for return loss,

impedance bandwidth can be calculated easily.

The second part, which is more complex, is the measurement of axial ratio bandwidth and

beamwidth. This involves the usage of Compact Antenna Test Range. The general steps

include measurements of the near radiation field of the antenna first, and then an assistant

software will transfer the near field pattern into far field pattern, which is the required

one. With these data, axial ratio bandwidth and beamwidth can be calculated.

37
 Section 5

EXPERIMENTAL RESULTS

AND COMPARISONS

In this section, the simulation results and experimental results will be compared for the

finalized antenna, which is also shown in figure 4.1.

One thing to note is that from table 3.3, the width of the bottom part of the U-slot is

actually 0.9mm. Due to fabrication restriction however, the smaller slot that can be cut is

38
only 1mm. Therefore, this may produce some influence over the actual performance of

the antenna compared with the simulated one.

5.1 Impedance bandwidth

The following figure shows the trend of return loss when the frequency increases from

2.02GHz to 2.73GHz obtained directly from Network Analyzer.

Figure 5.1 – Return loss S(1,1)versus frequency from network analyzer

For detailed comparison, numerical data are also obtained. Combining them with

simulation results, figure 5.2 is plotted with both results in the same graph for easier

comparison.

39
 Figure 5.2 – Simulation results and experimental results for S(1,1)

In the figure, the black line represents experimental result and the red line is simulation

result. Generally the experimental result matches well with the simulation result; they are

however, not perfectly the same. The impedance bandwidth from the experiment actually

shifts a little bit to the left. The exact impedance bandwidth from the experiment is 5.2%,

which is slightly larger than the simulation value (4.723%).

The possible reasons of causing this difference are:

40
(a) The dimensions of the U-slot during fabrication process are not exactly the same as

that in simulation. Besides the limit in width stated above, the corners of the U-slot are

not so sharp and follow a curve.

(b) There are some imperfections on the radiating plate of the antenna. For example, the

probe feed does not end exactly at the height of the radiating plate, but actually goes

1mm further.

(c) The simulation environment is ideal without interference, when using network

analyzer however, lots of factors should be taken in account, for instance, the topology of

the coaxial cable and electromagnetic wave in the environment.

(d) The simulation software IE3D adopts full wave analysis model to approximate the

antenna performance, so although it is persuasive, it is not perfectly identical to the exact

performance of the antenna because it also uses approximation.

5.2 Axial ratio bandwidth

41
The following figure shows the comparison between the actual and simulated axial ratio

when frequency changes.

Figure 5.3 – Simulation results and experimental results for axial ratio

With the red line representing simulation results and black line indicating experimental

results, figure 5.3 shows the axial ratio band actually shifts to the left like impedance

bandwidth. Calculation gives the experimental ARBW as 2.4%, which is smaller than the

simulation ARBW 2.9%. It is acceptable in normal telecommunication although it suffers

from a large reduction compared with simulation.

The reasons for this discrepancy are similar to those stated in 5.1, with some supplements

as follows.

42
The axial ratio actually depends on the radiated field of the antenna. Although in the

room of measurement, the space can be said to be free of external electromagnetic field,

the coaxial cable which feeds to the antenna actually produces radiation. And this will

influence the overall radiation pattern of the antenna.

5.3 Beamwidth

The radiation pattern of the antenna is shown in the following figure with red line

representing experimental results and blue line representing simulation results at

2.26GHz.

270˚

Figure 5.4 – Simulation results and experimental results for radiation pattern

Calculation gives experimental beamwidth of about 116˚ which is approximately the

same as simulated 120˚. The explanations for this difference are similar to that in part 5.3

as they all deal with radiation of the antenna.

43
 Section 6

44
 CONCLUSION AND

FUTURE DEVELOPMENT

To conclude, this project is to design a circularly polarized antenna with broadband

property and wide beamwidth. After mainly three stages of design process, a truncated

corner patch antenna with unbalanced U-slot surrounded with circular walls comes out.

The experimental performance of this antenna generally agrees with the simulation,

verifying the theories and methods in designing antenna.

For myself, over the one year of study, I have learnt a lot about antenna – from basic

theories to advanced design techniques. I have read a lot of papers regarding antenna

design, whether the focus is similar to the title of my project or not, and have done lots of

simulations during the design process. I can say I gained a lot by doing this project.

Some future development of the antenna may include:

45
(a) Reducing the ground plane of the antenna

(b) Enlarge the ARBW using Electromagnetic Bandgap Structure

For the method stated in (b), I have actually spent lots of time on it by placing small grid

over the ground plane, the effect however, was not so clear. But EBG structure is an

effective way to broaden the ARBW for many antenna types and therefore it is highly

possible that such structure will also be useful for my case.

Section 7
46
 REFERENCE

[1] http://www.k6if.com/arrl10_3.html

47
48