Design of circular polarized dual band patch antenna
Thomas Edling
Påbyggnadsprogrammet till civilingenjörsexamen i elektroteknik Master Programme in Electrical Engineering
Teknisk- naturvetenskaplig fakultet UTH-enheten
Besöksadress:
Ångströmlaboratoriet Lägerhyddsvägen 1 Hus 4, Plan 0
Postadress:
Box 536 751 21 Uppsala
Telefon:
018 – 471 30 03
Telefax:
018 – 471 30 00
Hemsida:
http://www.teknat.uu.se/student
Thomas Edling
At the moment Swedish Transport Administration uses a monitor system that can detect urgent errors as warm ball-bearings and flat wheels etc. with stationary detectors. To avoid these errors Swedish Transport Administration, UPWIS AB and Uppsala University work with a system that will continuously monitoring the train to detect the errors as fast as possible. This will save money in the future for Swedish Transport Administration and all other partners that use the rails.
Swedish Transport Administration has already RFID readers beside the rail to detect trains position. The new monitoring system will use these readers and send data from the monitoring system via these readers to a database.
The aim of this thesis work is to design and build a RFID antenna to send data from the monitoring system to the RFID readers. The antenna should be a circular polarized and it needs to manage the harsh environment on the train.
This thesis work started with a theoretical study which investigated four common antenna types (dipole, loop, PIFA and patch/microstrip) to evaluate which antenna type that is the best solution for this application. It was decided to design a patch antenna from the theoretical study since it fulfils all the requirements for the antenna.
Simulations and tests shows that the antenna is circular polarized and have a maximum reading distance of 5 m for 868 MHz. For 2.45 GHz it is linear polarized and has a reading distance of at least 10 m. With other hardware settings the antenna will have longer reading distance at 2.45 GHz.
When all parts of the test bed was finished the test bed was mounted on the measurement wagon. The final test shows that the antenna fulfils the task. The antenna transmitted the data from the sensor boxes to the RFID readers.
The report suggests future work to minimize the reading distance and size for the antenna. These are: transfer sensor data to RFID tag by “multi hop”, hardware improvement for instance antenna diversity and using another substrate (higher dielectric constant).
ISSN: 1654-7616, UPTEC E11008 Examinator: Nora Masszi
Ämnesgranskare: Anders Rydberg
Handledare: Mathias Grudèn, Magnus Jobs
varma kullager och platta hjul m.m. med hjälp av stationära detektorer. För att kunna
förhindra dessa fel arbetar Trafikverket, UPWIS AB och Uppsala universitet med att ta fram ett system som kontinuerligt ska övervaka tågen för att så fort som möjligt Detta ska
förhoppningsvis också spara en hel del pengar i framtiden för trafikverket och alla partner som använder rälsen.
Trafikverket har redan RFID läsare vid sidan av rälsen för att kunna detektera tågens position. Det nya övervakningssystemet använder dessa befintliga läsare och skickar data från övervakningssystemet via dessa läsare till en databas.
Målet med detta examensarbete är att designa och bygga en RFID antenn för att skicka data från övervakningssystemet till RFID läsarna. Antennen ska vara cirkulär polariserad antenn och kunna utsättas för den extrema miljö som den kommer befinna sig i.
Examensarbetet började med en teoretisk litteraturstudie som undersökte fyra vanliga antenntyper (dipol, loop, PIFA, patch/microstrip) för att utvärdera vilken antenntyp som skulle vara bäst lämpad för applikationen. Efter att litteratur studien gjorts bestämdes att en patchantenn var ett bra alternativ eftersom denna antenn typ kunde uppfylla alla krav som ställdes på antennen.
Simuleringarna och tester visar att antennen är cirkulär polariserad med ett läsaravstånd på 5 m for 868 MHz. För 2.45 GHz är den linjärpolariserad med ett läsaravstånd på minst 10 m.
Med andra hårdvaruinställningar kommer antennen ha ett längre läsaravstånd för 2.45 GHz.
När testbädden var klar och monterad testades systemet på tåget. Testet visade att RFID antennen uppfyllde uppgiften d.v.s. överförde data från sensornoden till RFID läsaren.
Rapporten tar även upp förslag på framtida arbete för att minska läsaravståndet för antennen och storleken på antennen. Dessa är användandet av ett annat substat
(högre εr), överföra sensordata till RFID taggen via ”multi hop” och hårdvaruförbättringar (antenndiversitet).
came.
I am grateful for all help by my subject inspector Prof. Anders Rydberg and my supervisors Mathias Grudén and Magnus Jobs at the Department of Solid State Electronics at Uppsala University. I am also grateful for all work Kjell Brunberg and Erik Jansson at UPWIS AB have done with this project these 5 month.
I would also thanks Ulf Hellström at Swedish Transport Administration and Infranord for all help.
Finally I would thanks my team mates (thesis worker) Malkom Hinnemo, Filip Zherdev and Nils Edvinsson with all help and support that have been needed during the thesis work with measurement etc.
I wish also all good luck with the project in the future. I think in the end the final version of the system will be very good and useful for Swedish Transport Administration.
List of figures………..…..i
List of tables………...….…iii
Chapter 1 ... 1
1 Introduction ... 1
1.1 Aim of Thesis Work ... 1
1.1.1 Antenna specification ... 1
1.2 Radio frequency identification ... 2
1.3 Structure of the project ... 3
1.4 Outlines for the thesis ... 3
Chapter 2 ... 4
2 Theory and concepts for patch antennas ... 4
2.1 Patch antenna design ... 4
2.1.1 Structure ... 4
2.1.2 Calculations for the patch antenna dimension ... 5
2.1.2.1 Length of antenna ... 5
2.1.2.2 Width of the patch antenna ... 7
2.1.3 Feed techniques ... 7
2.1.3.1 Microstrip Line Feeding ... 8
2.1.3.2 Coaxial Feeding ... 9
2.1.3.3 Coaxial probe with capacitive feed ...10
2.1.3.4 The aperture-coupled patch ...11
2.2 Dual band techniques ...12
2.2.1 Using higher modes ...12
2.3 Short-circuited patch ...14
2.4 Substrate ...15
2.4.1 Dielectric substrates ...16
2.5 Polarization ...17
2.5.1 Linear polarization ...17
2.5.2 Elliptical polarization ...19
2.5.3 Circular polarization ...20
2.5.3.1 Techniques for circular polarization ...21
2.5.4 Axial ratio ...22
2.6 Quality factor (Q-factor) ...22
2.7 Bandwidth ...23
2.7.1 Techniques for wider bandwidth ...24
2.8 Return Loss, S11 parameter ...25
2.9 Ground plane size effects ...26
2.10 Eddy current ...27
2.11 Metallic environments effect on antennas...28
2.12 Ferrite shielding ...28
3.1 Microstrip and Patch antenna...30
3.1.1 Advantages ...30
3.1.2 Disadvantages ...30
3.1.3 Tradeoffs ...31
3.2 Planar inverted f antenna ...31
3.2.1 Advantages ...33
3.2.2 Disadvantages ...33
3.2.3 Techniques for wider bandwidth ...33
3.2.4 Techniques for reducing the physical size ...33
3.3 Loop antenna ...34
3.3.1 Advantages ...34
3.3.2 Disadvantages ...34
3.4 Dipole antenna ...35
3.4.1 Advantages and disadvantages ...35
3.5 Decision for antenna type ...36
Chapter 4 ... 37
4 Design of dual band patch antenna ... 37
4.1 Design procedure ...37
4.2 Antenna design ...38
4.2.1 Starting and simulation stage ...38
4.2.2 Minimizing stage ...39
4.2.3 Matching stage to final stage ...40
4.2.3.1 Design errors ...40
4.2.4 Layout...42
Chapter 5 ... 46
5 Simulation results and tests ... 46
5.1 General simulations settings in CST ...46
5.2 General about simulations results ...46
5.3 Simulations result ...48
5.3.1 S11 parameter ...48
5.3.2 Smith chart ...51
5.3.3 Far field pattern ...52
5.3.4 Axial ratio ...54
5.4 Measure antenna with Network Analyzer ...56
5.5 Outside test ...57
5.6 Final set up on train ...60
6 Final words ... 633
6.1 Future work ...63
6.1.1 Update hardware ...633
6.1.2 Smaller RFID antenna ...633
6.1.3 Multi-hop system ...655
6.2 Conclusions...66
References ... 68
i
Figure 1.1: IEEE 802.15.4 channel selection ... 2
Figure 2.1: Structure of a rectangular patch antenna ... 4
Figure 2.2: Common forms of patch layer ... 4
Figure 2.3: Fringing field ... 5
Figure 2.4: Microstrip line feed ... 8
Figure 2.5: Coaxial feeding method ... 9
Figure 2.6: Coaxial probe with capacitive feed method ... 10
Figure 2.7: The aperture-coupled patch ... 11
Figure 2.8: Propagation of TEM, TE, TM waves ... 12
Figure 2.9: Geometry for field configuration ... 13
Figure 2.10: Electric field ... 14
Figure 2.11: Short-circuited techniques ... 15
Figure 2.12: Linear polarization ... 17
Figure 2.13: Horizontal, vertical linear polarization ... 18
Figure 2.14: Elliptically polarized wave ... 19
Figure 2.15: Circular polarization ... 20
Figure 2.16: Circular polarization (RHCP, LHCP) ... 20
Figure 2.17: Dual feed ... 21
Figure 2.18: Single feed ... 21
Figure 2.19: Axial ratio ... 22
Figure 2.20: Q factor... 23
Figure 2.21: 2 port network. ... 25
ii
Figure 3.1: PIFA ... 32
Figure 3.2: Patch layer for PIFA ... 32
Figure 3.3: Loop antenna ... 34
Figure 3.4: Dipole structure ... 35
Figure 4.1: Design procedure ... 37
Figure 4.2: Simulation error ... 41
Figure 4.3: Top view on feed ... 42
Figure 4.4: Top view on ground ... 42
Figure 4.5: Top view on patch... 43
Figure 4.6: Side view (long side) ... 43
Figure 4.7: Side view (Short side) ... 43
Figure 4.8: Built antenna, top view ... 43
Figure 5.1: S parameter for 868 MHz ... 48 11 Figure 5.2: parameter for 2.45 GHz ... 49
Figure 5.3: parameter for 868 MHz and 2.45 GHz ... 50
Figure 5.4: Smith chart for 868 MHz ... 51
Figure 5.5: Smith chart for 2.45 GHz ... 51
Figure 5.6: Far field pattern for 868 MHz ... 52
Figure 5.7: Far field pattern (left side) for 2.45 GHz ... 53
Figure 5.8: Far field pattern (right side) for 2.45 GHz ... 53
Figure 5.9 a-b: Axial ratio for 836 and 868 MHz. ... 54
Figure 5.10: Axial ratio 2.45 GHz (left side) ... 54 S11
S11
iii
Figure 5.12 a-c: S11 parameter from Network analyzer for 868 MHz and 2.45 GHz. ... 56
Figure 5.13: Set up for the outside test ... 57
Figure 5.14: RFID reader ... 59
Figure 5.15: Set up for final test... 60
Figure 5.16: Sensor node box on train ... 61
Figure 5.17: Measure wagon for the final test ... 62
Figure 6.1: Small RFID antenna ... 64
Figure 6.2: Set up for multi hop-system ... 64
List of tables Table 1: Tradeoffs patch antenna……….……….……...31
Table 2: Parameter value for built antenna………...44
Table 3: Antenna position, outside test………...………..…...58
1
Chapter 1
1 Introduction
Todays monitoring of railways wagons are used to discover warm ball-bearings, locked brakes, flat wheels and other different problems that can occurs on the train. At the moment this is done by stationary detectors. It means that it is only detecting urgent errors which can cause unplanned delays in the traffic. Therefore it is very desirable to continuously monitoring the trains and store the information from the sensors on the trains in a database.
Together with Swedish Transport Administration, UPWIS AB and Uppsala University the goal with the project during summer and autumn 2011 is to build and install a test bed based on wireless sensors on the railway wagons. Swedish Transport
Administration, UPWIS AB and Uppsala University are together a part of the research center Wireless Uppsala VINN Excellence Center for Wireless Sensor Networks (WISENET)
1.1 Aim of Thesis Work
During autumn 2011 a test bed will be finished if all goes by the schedule. The aim with this thesis work is to; build an antenna for this test bed that will transfer data as
temperature and vibration from different types of wireless sensors to readers beside the rail.
1.1.1 Antenna specification
Environmental conditions: In a metallic environment on railway wagons.
Polarization at 868 MHz: Circular polarization
Read range at 868 MHz: The RFID antenna has a minimum vertical and horizontal1 read range. The minimum vertical read range is 3 m and the minimum horizontal read range is between 1.6-3.8 m depending on which settings are used for the hardware. [1]
1 Vertical and horizontal direction, see figure 5.14.
2
Polarization 2.45 GHz: Depending on transmitting antenna Read range 2.45 GHz: Minimum range 25 m
Physical size: As small as possible, no maximum size
Frequency band: The 868 MHz antenna uses the ISO 18000-6 type C standard,
860-960 MHz. The main goal is not to cover the whole frequency band, only the part of the band that is used in Europe, i.e. 865.7-867.6 MHz.
The 2.45 GHz antenna operates with the IEEE 802.15.4 standard, 2.4-2.4835 GHz.
Figure 1.1: IEEE 802.15.4 channel selection [2]
The IEEE 802.15.4 frequency band is divided into 16 channels. Figure 1.1 shows how the channels are divided. Each channels center starts at 2.400 0.005 ,
16 ,..., 3 , 2 , 1
=
n and are 2 MHz wide. [2]
1.2 Radio frequency identification
Radio frequency identification (RFID) belongs to the technology automatic
identification and data capture (AIDC). This technology collects data automatically from tags so it can be used [2]. RFID uses radio waves to transfer data from the electrical tags to the reader. The technology is used in many applications as asset tracking, manufacturing, supply chain management, payment systems, security and access control. The tags can be divided into two groups, active and passive. The active tags have their own power supply and the passive activates by induction from the reader’s antenna. [3]
3
Phase 1: The project begins with to investigate the advantages and disadvantages of few different common types of antennas. After that, determine which antenna type that could be a good choice for the project.
Phase 2: Find solutions how to build the antenna so it meets the antenna specification.
Phase 3: Build the antenna.
Phase 4: Test the antenna.
Phase 5: Implement the antenna on the train.
1.4 Outlines for the thesis
Chapter 2 presents useful theory for the design of the patch antenna. In chapter 3 different types of antennas is investigated to decide which antenna type that is a good solution for this project. Chapter 4 shows the design procedure of the antenna. Chapter 5 presents test results from different tests of the built antenna. Chapter 6 summarizes the thesis work and gives recommendation of future work
4
2 Theory and concepts for patch antennas
2.1 Patch antenna design
2.1.1 Structure
The patch antenna belongs to the class resonant antennas. For a rectangular patch antenna see figure 2.1. It is resonant when the length, L is around half multiples of the resonant frequency. The patch antenna consists in general of three major layer, ground plane, substrate and patch. [4]
Figure 2.1: Structure of a rectangular patch antenna [4]
Figure 2.2: Common forms of patch layer [4]
5
common forms of patch layer. Circular and triangular is used often instead of rectangular because it is easier to derive the mathematical expression for the model (antenna). [4]
2.1.2 Calculations for the patch antenna dimension
To design one simple patch antenna following parameters needs to be calculated:
length, width and eventual feed line for microstrip antenna.
2.1.2.1 Length of antenna
Figure 2.3: Fringing field [4]
To calculate the length of the patch antenna the fringing fields that occurs needs to take into account. The fringing field occurs at the ends of the patch. The electric field does not end abruptly at the edges and therefore create the “fringing fields”. These fields can be represented as two radiation slots which means that the patch looks electrically larger than the physical size. Because of that the calculated length need to be extended with the fringing factor ∆ so the antenna design is for patch with /2 and no fringing. [4]
6
( ) ( )2.1
2 2
1
0
0 eff
r L L
f = ε µ + ∆ ε
where ε0 is the permittivity in vacuum,µ0 is the permeability in vacuum, Δ is the fringing factor and eeff is the effective electric constant which take the fringing field outside the patch into account [5].
Effective electric constant give by formula [4]
) 2 . 2 ( 1
/ , 12 2 1
1 2
1 + >
−
+ +
= W h
W
r h
r eff
ε ε ε
Fringing factor gives by formula [4]
) 3 . 2 8 (
. 0 /
264 . 0 / 258 . 0
3 . 412 0
.
0
+ +
−
= +
∆ W h
h h W
L
eff eff
ε ε
To optimize the length and resonant frequency with formula (2.1), a praxis value for L is used. [6]
) 5 . 2 (
) 4 . 2 ( 49
. 0
~ 48 . 0
r r g
g g
f c L
λ ε
λ λ
=
=
where c is the velocity of light in vacuum.
7
2.1.2.2 Width of the patch antenna
The width of the patch gives by formula [4]
) 6 . 2 1 (
2 2
1
0
0 +
=
r r
f
W µ ε ε
It is recommended that the width of the patch is in following interval [7]
) 7 . 2 ( 2L
W L< <
2.1.3 Feed techniques
There are many methods to feed the patch and all have their advantages and
disadvantages. These feeding methods can be classified into two groups, contacting and non-contacting. For the contacting methods, the patch antenna feeds directly to the patch and for the non-conducting method electromagnetic field coupling is used to transfer the power to the patch.
8
2.1.3.1 Microstrip Line Feeding
The microstrip line consists of a conducting strip connected to the patch. The microstrip line has often the same thickness as the patch but the width is smaller. To obtain good impedance matching an inset cut (x0) can be made. The length of the inset controls the impedance matching [4]. Figure 2.4 shows how to use microstrip as feed technique.
Figure 2.4: Microstrip line feed [4]
Advantages with microstrip feed are [4]:
• One of the easiest methods to fabricate.
• Easy to match by controlling inset length.
Disadvantages with microstrip feed are [4]:
• Give undesirable cross polarization effects.
• Make the patch larger.
• Bandwidth decreases when the thickness of the substrate increases.
9
2.1.3.2 Coaxial Feeding
The coaxial feed method is one of the most common feed techniques. The inner conductor of the coaxial goes through the substrate from ground to the patch and the outer conductor are connected to the ground plane [4]. Figure 2.5 shows how to use coaxial probe as feed technique.
Figure 2.5: Coaxial feeding method [4]
Advantages with coaxial probe feed are [4]:
• Easy to fabricate.
• Easy to match because the feed position can be placed anywhere to the patch to get impedance matching.
Disadvantages with coaxial probe feed are [4]:
• Narrow bandwidth.
• For thicker substrates, the increased probe length makes the input impedance more inductive which lead to matching problem.
• Difficult to model specially for thick substrate.
10
2.1.3.3 Coaxial probe with capacitive feed
The difference between the usual coaxial feed and this method is that the inner
conductor of the coaxial goes not the whole way up to the patch and the end of the inner conductor is connected to a circular plate. If a regular probe were used, a larger
inductance would be introduced, which results in impedance mismatch. To cancel the inductance a reactance need to be added. This feeding method with the capacitive disk does that [8]. Figure 2.6 shows how to use coaxial probe with capacitive feed as feed technique.
Figure 2.6: Coaxial probe with capacitive feed method [8]
Advantages with coaxial probe feed are [9]:
• Wide bandwidth.
• The capacitive disk cancels the inductive impedance of the probe.
Disadvantages with coaxial probe feed are [10]:
• May reduce efficiency.
• Can be difficult to design, depending on substrate. Not difficult with air.
11
2.1.3.4 The aperture-coupled patch
Below the patch are there two layers of substrate. The substrate layers are separated with a ground plane. A microstrip line is placed below the lower substrate layer. The energy is coupled to the patch from the micrstrip line by a slot in the patch. Figure 2.7 shows how to use the aperture-coupled patch as feeding technique. [4]
Figure 2.7: The aperture-coupled patch [4]
Advantages with the aperture-coupled patch feed are [4]:
• Many parameters to choose between to match the antenna as height of substrate, width, length and position of the slot.
• Purifier polarization since the feed is isolated by the ground plane between the substrates.
Disadvantages with the aperture-coupled patch feed are [4]:
• Difficult to fabricate.
• Narrow bandwidth.
12
2.2 Dual band techniques
2.2.1 Using higher modes
The region between the patch and the ground plane for a rectangular patch (see figure 2.1) can be viewed as a rectangular box with electric walls above and below it and magnetic walls at the sides. The whole system can be treated as a rectangular waveguide [4]. There are different types of waveguides and in them can three different types of electromagnetic waves propagate, TEM, TE and TM [11], [12].
Transverse Electro Magnetic (TEM) wave: The electric and magnetic field are perpendicular to each other and both are transverse to the direction of propagation ( y ).
There is no component for the electric and magnetic field in direction of propagation (Ey =Hy =0). TEM wave exist in waveguide that consist of two conductors, as coaxial.
Transverse electric (TE) wave: The electric and magnetic field are perpendicular to each other and both are transverse to the direction of propagation ( y ) and in the propagation direction Hy ≠0 andEy =0. TE wave exist in waveguides that consist of one conductor.
Transverse magnetic (TM) wave: The electric and magnetic field are perpendicular to each other and both are transverse to the direction of propagation ( y ) and in the
direction of propagation Ey ≠0andHy =0.
Figure 2.8: Propagation of TEM, TE, TM waves
13
Figure 2.8 shows the direction of electric and magnetic field for TE, TM, TEM waves when z-axis is the direction of propagation for the TEM, TE and TM waves in the waveguide. For patch antennas only TM waves can propagate in the dielectric rectangular waveguide.
Figure 2.9: Geometry for field configuration [4]. TM wave propagates with E-field in y and z, y-direction and H-field in x-direction.
From the expression for the field configuration, the resonant frequencies for the waveguide can be calculated. Figure 2.9 shows the geometry for the cavity that is used for the field configuration. The resonant frequencies for the cavity are given by [4]
( ) (2.8)
2
1 2 2 2
+
+
=
W p L
n h
fr mnp mπ π π
µε π
In the x-axis the electric field varies negligibly [13], [4] and formula (2.8) can be approximated to following resonant frequency
( ) (2.9)
2
1 2 2
+
=
W p L
fr np nπ π
µε π
14
It is common that microstrip and patch antennas are short-circuited. The techniques that are used are shorting pin in the patch or a shorting wall along a line at the end of the patch. In applications there a small size antenna are needed a shorting wall is a good alternative. For a half-wave rectangular patch the electric field distribution is given by
) /
0cos( x L
E π .
Figure 2.10: Electric field TM010x [4]
Figure 2.10 shows that the electric field (TM010x ) has a maximum at the both edges of the patch and zero in the middle. There is 180o phase between the fields maximum. Since the electric field is zero in the middle (planex=L/2) an electric wall can be placed there. The shorting wall will not change the designed resonant frequency for the half- wave rectangular patch [10]. This geometry of the patch antenna is called quarter wave patch because the distance between the radiation edge and the shorting wall isλ/4. The largest difference between a half wave patch and a quarter wave patch is that the quarter wave patch has one radiation edge and half wave patch has two. Since the structure is different between a half wave patch and a quarter wave patch it gives changes for the antenna characteristics. This is few of them [5].
1. The quarter wave patch has a broader E plane pattern since the half-patch has two radiation edges.
2. The quarter wave patch has lower gain since it only has one radiation edge, half wave patch has two.
3. The ration between half wave patch and quarter wave patch is two for the radiation conductance, one half for the radiation resistance, εr ≠1 and two for the stored energy.
15
Figure 2.11: Short-circuited techniques [5], [10]
Figure 2.11 show different ways to add shorting pins and shorting wall. It can be done by one or few shorting pins. The shorting wall can also consist of shorting pins.
2.4 Substrate
Substrate materials play an essential role for the patch antenna design. The substrate have many properties that should be considered: the dielectric constant, loss tangent, their variation with temperature and frequency, homogeneity, isotropic, thermal coefficient and temperature range, dimensional stability with processing and temperature, humidity and aging, and thickness uniformity of the substrate [3]. One properties that substrate has is the permittivity. The permittivity is associated with how much electrical charge a material (substrate) can store in a given volume. The
permittivity(ε is complex and has one real part ) (ε') and one imaginary part(ε'') [14].
) 10 . 2
'' (
' ε
ε ε = − j
The loss tangent (tan ) measure the amount of electrical energy converted to heat in δ the dielectric and accounts for the power losses in passive device such the transmission line or patch antenna and defined as the ratio between the real part and imaginary part of the complex permittivity [10], [14].
) 11 . 2 ( )
tan( ''' ε δ = ε
16
0
[15].
) 12 . 2 (
0 '
ε εr = ε
Since the speed of propagation in given medium is [15].
) 13 . 2 ( ]
/ 1 [
1 0
0 0
s c m
c
r r r
rµ µ ε µ
ε ε
εµ = =
=
The dielectric constant affects the speed it will also affect the wavelength and frequency [15].
) 14 . 2 ( ]
0 [
m f
c f c
εr
λ = =
2.4.1 Dielectric substrates
One common type of substrate is dielectric substrates. The substrate is used to fulfill two different factors, mechanical support for the structure and determines the electrical characteristics of the circuit or antenna. [10]
Mechanical properties
• Mechanical strength, for example vibration resistance and shape stability.
• Small dilatation factor.
Electrical properties
• Relative permittivityεr, which determines the miniaturization factor. If all other parameters are fix the size of the circuit is proportional to 1/ εr .
• Small dielectric losses. Should havetan <δ 0.001.
17
The polarization of an antenna in a given direction is defined as “the polarization of the wave transmitted (radiated) by the antenna” [4].
2.5.1 Linear polarization
An antenna has linear polarization if the field vector (electric or magnetic) for the transmitted wave of the antenna only has one component or two orthogonal linear component that are in time phase or 180o (or multiples of 180o) out of phase [4].
Figure 2.12: Linear polarization [16]
18
Figure 2.13: Linear polarization [17]
Figure 2.12 shows that the electric field propagates in one plane and the magnetic field propagates in the plane orthogonal to that. Figure 2.13 shows that the linear polarized wave can propagate in two different planes. When it propagates in the x-y plane it is horizontal linear polarization and in the x-z plane it is vertical linear polarization.
19
An antenna has elliptical polarization if the field vector (electric or magnetic) for the transmitted wave of the antenna has two orthogonal linear components that can have the same or different magnitude. When these components have different magnitude the time-phase difference between the components can not be 0o or multiples of 180o because it will then be linear polarization. When the two components have the same magnitude the time-phase difference between the two components must not be odd multiples of 90obecause it will be circular polarization. [4]
Figure 2.14: Elliptically polarized wave [16]
Figure 2.14 shows how the elliptically polarized wave propagates. The electric field propagates in two planes at the same time with different amplitude.
20
An antenna has circular polarization if the field vector (electric or magnetic) for the transmitted wave of the antenna has two orthogonal linear components with the same magnitude and the time-phase difference between this is odd multiples of 90o.
There are two types of circular polarization. If the field rotation is clockwise then right- hand circular polarized (RHCP) or if the field rotation is counterclockwise then left hand circular polarized (LHCP) [18].
Figure 2.15: Circular polarization [18]
Figure 2.16: Circular polarization [18]
Figure 2.15 shows how the circular polarized wave propagates. The electric field propagates in two planes at the same time with equal amplitude. Figure 2.16 shows that if the electric field rotates clockwise (RHCP) and counterclockwise (LHCP).
21
There are many ways to get circular polarization for the patch antennas. Figure 2.17- 2.18 shows some common ways to get left-and right handed circular polarization. They can be divided into two groups, dual feed and single feed patches.
Figure 2.17: Dual feed [10]
Figure 2.18: Single feed [10]
The dual feed patch antennas feeds with equal amplitude but with 90o phase difference.
The single feed patch antennas using truncated corner, truncated circle, corner feed or slots in the patch. [8], [10].
22
The axial ratio ( ) B AR A
Ο
= Ο is the ratio between the length of the major axis ( AΟ ) and the minor axis ( BΟ )of the E-field, (ΟA≥ΟB). The E-field consists of an x, y
component(Ex,Ey). [4]
Figure 2.19 Axial ration: A electric field created by the two electric field component Ex and Ey. [4]
For a linear polarized wave the AR =∞(dB>> 0) since the electric field has one component and thenΟB=0. For a circular polarized wave the major and minor axes have equal length (ΟA=ΟB) which givesAR =1 dB(0 ). For an elliptical polarized wave the major and minor axes have different length which gives1< AR<∞. [4]
2.6 Quality factor (Q-factor)
Q factor is a parameter that describes how much power that transform as losses in the system. A high Q factor indicates a lower rate of energy loss relative to the stored energy. [19]
) 15 . 2 (
2 0
1 2 0 0
Bandwidth f f
f Q f Loss or Power
stored Energy f
Q =
= −
= π
23
Figure 2.20: Figure shows how f0, f1 and f2
are defined for Q factor calculations. [19]
Figure 2.20 shows the amplitude in function of the frequency for the antenna. For a given center frequency( )f0 , f1 and f2 is the frequency when the center frequency has dropped 3 dB from the maximum value. A low Q factor gives a broad band (wide) bandwidth or a high Q factor gives a narrow band (small) bandwidth.
The quality factor can also be expressed as a function depending of the substrate thickness h and the dielectric constantεr [8].
) 16 . 2 4f0h (
Q c εr
=
Q decrease when εr decrease or h increase.
2.7 Bandwidth
The bandwidth is defined as “The range of usable frequencies within which the
performance of the antenna, with respect to some characteristics, conforms to a specified standard” [20].
There are many ways to calculate the bandwidth, but for narrow band antennas the fractional bandwidth (Bfrac) is common [4].
24
) 18 . 2 ( 100
*
0 1 2
f f Bfrac f −
=
where f2 = upper frequency, f1 = lower frequency and f0 = center frequency.
Formula 2.19 shows how the bandwidth depends on different parameters for a rectangular patch [5]
) 19 . 2 ( 1 ,
77 .
3 2 λ
ε λ
ε − <<
= h h
L B W
r r
where εris the permittivity, Wis the width of the patch, L is the length of the patch, h is height of substrate, λis the wavelength
The bandwidth can also be calculated from formula
) 20 . 2 4 02 (
0
c r
h f Q B f
= ε
≈
2.7.1 Techniques for wider bandwidth
There are many techniques for higher bandwidth. This is few of the most common [10].
• Thicker substrate
• Bigger antenna, in general. When the antenna is electrically small the bandwidth decrease [21].
• Lower permittivity
• Different feeding techniques gives different bandwidth. Choose feeding technique that gives wider bandwidth.
25
If the antenna has two ports, input and output (two ports network), the incident power to the antenna can be reflected, radiated or absorb as losses.
As figure 2.21 shows,
) 21 . 2
loss (
rad r
i P P P
P = + +
Figure 2.21: 2 port networks
Many antennas are designed with low loss materials which mean that most of the power that is not reflected is radiated. [22].
When the characteristic impedance of the transmission lines are different from the impedance at the input port at the antenna, this lead to losses of the incident power from the 2 port network system and part of the incident power will be reflected back through the transmission line.
The return loss or reflection loss measures the effectiveness of the power delivery from the transmission line to the antenna [23]. The return loss is identified by the
S11 parameter from two port network theory. The return loss (RL) is defined as [24]
) 22 . 2 ( log
20 log
20 log
10 )
( 10 10
S L
S L i
r r
i
Z Z
Z Z E
E P
dB P
RL +
− −
=
−
=
=
where Pi is the Incident power [watt], P is the reflected power [watt], r Eiis the incident electric field, Eris the reflected electric field, ZL is the load impedance, ZS is the transmissions line characteristic impedance [23], [24].
From formula 2.22 the reflected power can be calculated for a two port network.
26
2.9 Ground plane size effects
It has been shown that the size of the ground plane affects the excitation efficiency of non-resonant modes. Therefore the thickness of the ground plane can be used to modify the mode excitation. For the patch antenna design with small ground plane it is not only to optimize the patch size but also the size and thickness of the ground plane. [10]
In many cases for antenna design one goal is to reduce the antenna size inclusive the ground plane. In many designs the ground plane is larger than the patch. There is a problem with finite ground plane. It can raise the diffraction of radiation from the edges of the ground plane which can lead to changes in the radiation pattern, radiation
conductance and resonant frequency. Studies have shown when the antenna size is equal to the ground plane size it result in higher resonant frequency which not an infinitely ground plane gives. If the ground plane width and length increase with the length
SEXT, the fractional change in resonant frequency is given by [5]
( / ), / 1 (2.24)
1 ln lim 1
0 0
0
>>
≈ +
∆
→ h W h forW h
f f
r r S r
k EXT ε λ
ε π
where ∆f is the changes in resonant frequency with the extensional ground plane.
When formula (2.24) is not satisfied for theSEXT, formula (2.25) is used instead.
) 25 . 2 240 (
W B h f
f
r =− εre
∆
where
( ) ( )
[ ] (2.26)
8 0
2 0 0
2 0 0 0
EXT
EXT Y k S
S k W J
B=k −
η
where J0, Y0 is the first and second order of the Bessel function.
Another common recommendation for ground plane size is [25]:
27
) 28 . 2 ( 6 antenna patch
ground h W
W = +
When the height of the antenna increase the size of the ground plane increase which gives a noticeable effect on the total size of the antenna for small high antennas, i.e. the ratio between Lpatch/hantenna or Wpatch/hantenna decrease.
2.10 Eddy current
“Eddy current” is current induced in conductors. If a large conductive metal plate is moved through a magnetic field which intersects perpendicularly to the sheet, the magnetic field will induce small "rings" of current which will actually create internal magnetic fields opposing the change. Figure 2.22 shows were the eddy current can be created for a coil over a conducting material. [26]
Figure 2.22: Eddy current [26]
28
Metallic environment can have huge consequences for the antenna characteristics (antennas without or with ground plane). The eddy currents and image theory are two explanations why antennas do not work so well in metallic environment.
Eddy currents: When a metal plate or similarly is placed near an antenna the magnetic field (from coil in figure 2.22) creates eddy current which occurs in the metal plate.
These current gives undesirable effects as absorption of power which leads to detuning of the antenna. This detuning will decrease the operating distance and quality factor [27].
Image theory: When a dipole is placed near a metal surface it will produce a image.
The image has current with opposite sign. The image produces an electric field which will cancel the electric field caused by the original source (dipole). The effect can be negligible by move the dipole away from the metal surface. This means that you can not place an antenna on a metal surface and expect it to radiate. Consequences of the image theory are that the antenna can get detuned (shift in different resonant frequency). The metal will also affect the impedance at the antenna since a capacitance will be
introduced and this will reflect a part of the energy back to the terminals because the antenna is miss matched [27], [28].
2.12 Ferrite shielding
To reduce the generated eddy currents ferrite material can be used as a shield between the antenna and the metal behind the antenna [27]. Ferrite materials have high electrical resistance, up to106Ω which decrease the effect of the eddy currents [29]. Electrical resistance is the resistance that materials prevent electrical current flow through for example a conductor. The electrical resistance increase when the conductor is longer or the area is smaller [30]. The eddy current reduce operating distance but a ferrite
shielding will not increase the operating distance above values achievable in non metallic environment. [27]
The ferrite is a substance that consists between a mix of iron oxide and oxides (or carbonates) with other related material, such as nickel, magnesium and zinc. Ferrites are used in many applications that involve magnet induction. [27]
29
By choosing ferrite materials there is a large section of materials depending which combination of iron oxide and carbonate which are used. Ferrite materials have a high Q value and resistivity. The ferrites can be shaped in many different forms and it is a good choice for low cost applications. [31]
Disadvantages with ferrite materials
In many applications ferrite materials is not a good choice since it is a tenuous material, especially pure ferrite. This lead also to problem when it will be used in machines. To solve this problem the pure ferrite is mixed with other materials that will change the materials properties. Mix ferrite with other materials can be both expensive and take time. Ferrite is also weak at thermal shocks. [32]
30
3 Design ideas for antenna
The first phase in the thesis work was to investigate what type of antenna that meets the requirements for the antenna. The investigated antenna types are microstrip, patch, PIFA, loop and dipole. These are all very common antenna types and are used in many applications.
3.1 Microstrip and Patch antenna
3.1.1 Advantages
Patch antennas can be made small and flat. One layer of 1.5 mm FR-4 is used in many applications which give them a low profile. The patch antennas can also be very light weighted. Since the patch antennas only consist of two metal surfaces and a substrate layer the patch antennas can be made at a low cost depending on the choice of substrate and feeding technique. FR-4 is relative a cheap substrate material. In many designs patch antennas is easy to manufacture. The structure of a patch antenna makes it possible to integrate the antenna with circuits, the ground plane is a useful property and it simple to create array antennas. [10]
3.1.2 Disadvantages
Since the patch antennas are often small it affect many of the antenna properties negatively. The patch antenna has often low efficiency and narrow bandwidth. The different types of feeding techniques can give surface waves. The antenna is known for their tolerance problem and it requires good quality of the substrate to got high
temperature tolerance. It can also be difficult to make arrays with high performance since it requires complex feed system. Design an antenna with good purity of the polarization can also be difficult to achieve. [10]
