Printed Circuit Board Dipole Antennas and Dipole Antenna Array Operating at 1.8GHz

Transcription

1 Declaration I declare that this thesis is my own unaided work, and hereby certify that unless stated, all work contained within this paper is my own to the best of my knowledge. This thesis is being submitted for the degree of Bachelor of Science in Engineering (Electrical Engineering) at the University of Cape Town. Tai-Lin Chen 13 October 2003 i

2 Acknowledgements I would like to thank my supervisor, Professor B J. Downing for his guidance and expertise throughout this thesis. I am grateful to the following people for their helpful advices and assistance: Professor M. Reineck for the advices that he has given me, especially on testing the antennas. Mr. P. Daniels for his advices on etching the printed circuit board antennas. Mr. M. Lennon for demonstrating to me on how to crimp BNC cables. My colleagues for their assistance in gathering information needed for this thesis. ii

3 Terms of Reference This thesis was commissioned by Professor B J. Downing on the 14 th of July 2003 in particular fulfillment of the degree in Electrical Engineering at the University of Cape Town. Professor Downing s specific instructions were: 1. To investigate and understand the basic principles and concepts involved in designing antennas and antenna arrays. 2. To determine the most suitable antenna for 1.8GHz commercial applications. 3. To design and fabricate the antennas and antenna array. 4. To test and analyze the antennas and antenna array built. 5. To draw conclusions based on the findings. 6. To make appropriate recommendations for improvements and future research. 7. To submit the thesis by 21 st October iii

4 Synopsis The two most widely used frequencies for radio communications around the world in recent years are the 900MHz and the 1.8GHz bands due to the global rise in the industry of cellular networks. Most wireless communication devices manufactured today are using monopole antennas owing to it simple structure, omnidirectivity and its small size. However, these factors limit its signal strength because of the low gain. This is particularly apparent in the 1.8GHz band. A solution for this problem lies within the printed circuit technology, for its compact and cost effective nature. This thesis involved determining the most suitable printed circuit board antenna to use, designing it, implementing a suitable antenna array and testing it for the purpose of 1.8GHz wireless application. The design project will include understanding the principle theories and properties of antenna, microstrip, balun (balanced to unbalanced transformer) and antenna array. These important factors were acknowledged and understood before design procedures took place. After studying different types of antennas, designing of PCB dipole antennas were proposed due to its appropriate nature, these being omnidirectional, compact and cost iv

5 effective. As a dipole antenna needs a balun to be compatible for connection to a co-axial cable, hence balun designs were also proposed. An important objective of this thesis was to provide a solution to the setback of having low antenna gain in mobile wireless devices. The proposed solution is to design a suitable antenna array to increase antenna gain, and at the same time retaining the omnidirectivity desired. A four elements corporate fed array was then chosen as the array design, as its simple configuration meant that omnidirectivity and manipulation of the arrays were possible if further testing were to be conducted for optimum antenna performance. The design procedure includes the following: Building of prototype antennas to assists in familiarizing with fabrication techniques, and eliminating erroneous designs. Design and fabrication of etched PCB dipole antennas. Design and fabrication of etched antenna array components Tests and analysis was carried out on etched products after it was fabricated. The tests conducted include Impedance matching of all components, and the power radiation of the antenna array and the array element employed. Conclusions were drawn from the test results, indicating that the antenna array built is satisfactory and have achieved the objective set for this thesis, which is to design and build a high gain omnidirectional printed circuit board dipole antenna array operating at 1.8GHz. Recommendations are then made on the improvements, future testing v

6 methods and possible application of the antenna array. Further research and testing on the current antenna array has also been proposed. vi

7 Contents DECLARATION...I ACKNOWLEDGEMENTS... II TERMS OF REFERENCE...III SYNOPSIS...IV CONTENTS... VII LIST OF FIGURES...XI LIST OF TABLES...XIII INTRODUCTION BACKGROUND Printed Circuit Board Antennas Antenna Array OBJECTIVES LIMITATIONS OF RESEARCH SOURCE OF INFORMATION PLAN OF DEVELOPMENT... 4 LITERATURE REVIEW... 6 vii

8 2.1 ANTENNA PROPERTIES Impedance Wavelength Gain Effective Area Directivity Efficiency Polarization PRINTED CIRCUIT BOARD ANTENNA Types of Microstrip Antenna Microstrip Line Discontinuity BALUN ANTENNA ARRAY Array Theory Array Architecture DESIGN CONSIDERATION PROPOSED DESIGN PROPOSED PCB DIPOLE RADIATOR DESIGN PROPOSED BALUN DESIGN PCB DIPOLE ANTENNA ARRAY DESIGN DESIGN AND FABRICATION METHODS DESIGN PROCEDURES BUILDING AND TESTING OF PROTOTYPE MICROSTRIP ANTENNAS viii

9 4.2.1 Prototype Designs Prototype Performance DESIGNS OF PCB ANTENNAS ARRAY DESIGN TESTS AND ANALYSIS OF PCB DIPOLE ANTENNAS AND ANTENNA ARRAY IMPEDANCE MATCHING OF ETCHED ANTENNAS S11 2 Test on PCB Dipole Antennas Error Analysis Frequency Correction on the 10mm Design ANTENNA ARRAY TEST AND ANALYSIS Array Element Performance Array Power Splitter Antenna Array Performance CONCLUSIONS AND RECOMMENDATIONS CONCLUSIONS Results of Array Elements Results of Array Power Splitter Results of Antenna Array RECOMMENDATIONS Improvements on the Antenna Array Antenna Testing Recommendations Possible Antenna Array Application Proposed Further Testing BIBLIOGRAPHY ix

10 APPENDIX A APPENDIX B APPENDIX C x

11 List of Figures Figure 2.1: Elliptically Polarized Electromagnetic Field Figure 2.2: Microstrip Line Figure 2.2.3: Discontinuity Equivalent Circuit Figure 2.4.1: Array Figure 2.4.2: Four Elements Corporate Fed Array Figure 3.1: Proposed Dipole Radiator Design Figure 3.2: Proposed Balun Designs Figure 3.3: Balun Designs Figure 3.4: Proposed Array Design Figure 4.1.1: Prototype Design A Figure 4.1.2: Prototype Design B Figure 4.1.3: Balun to Ground Plane Width Ratio 3: Figure 4.1.4: Prototype Design C Figure 4.1.5: S11 2 of Prototype Design A Figure 4.1.6: S11 2 of Prototype Design B Figure 4.1.7: S11 2 of Prototype Design C Figure 4.2.1: 8mm PCB Dipole Figure 4.2.2: 10mm PCB Dipole xi

12 Figure 4.2.3: 12mm PCB Dipole Figure 4.2.4: Wideband PCB Dipole Figure 4.3.1: Corporate Fed Array Power Splitter Figure 5.1.5: 10mm PCB Dipole after Frequency Correction Figure5.2.2: Polar Radiation Test Configuration Figure 5.2.3: Polar Radiation of an Array element Figure 5.2.4: Impedance Terminator Connections Figure 5.2.6: Antenna Array Figure 5.2.7: S11 2 of Antenna Array Figure 5.2.8: Polar Radiation of Antenna Array Figure B1: Photograph of Prototype Design A S11 2 Test Result Figure B2: Photograph of Prototype Design B S11 2 Test Result Figure B3: Photograph of Prototype Design C S11 2 Test Result Figure B4: Photograph of 8mm PCB Dipole S11 2 Test Result Figure B5: Photograph of 10mm PCB Dipole S11 2 Test Result Figure B6: Photograph of 12mm PCB Dipole S11 2 Test Result Figure B7: Photograph of Wideband PCB Dipole S11 2 Test Result Figure B8: Photograph of 1.8GHz PCB Dipole S11 2 Test Result Figure B9: Photograph of Array Power Splitter S11 2 Test Result Figure B10: Photograph of Antenna Array S11 2 Test Result xii

13 List of Tables Table 2.1: PCB Antenna Characteristics Table 4.1: Prototype S11 2 Analysis Table 5.1: PCB S11 2 Analysis xiii

14 Chapter 1 Introduction This thesis describes the design, construction and testing of printed circuit board dipole antennas and dipole antenna array operating at 1.8GHz. 1.1 Background In recent years the two most widely used frequencies for radio communications around the world are the 900MHz and the 1.8GHz bands due to the global rise in the industry of cellular networks. Yet as we progress through this rapidly growing market, one important factor that has immense potential to improve is the signal strength of the cellular devices that we carry. Most wireless communication devices that have been marketed today are using monopole antennas due to it simple structure, omnidirectivity and its small size. It is due to these factors that its signal strength is limited by the low gain of these antennas, this is especially apparent in the 1.8GHz band. [1] As these commercial devices evolve, the demands for better performance will rise, 1

15 yet these products must still keep its limitations within the boundaries of being easy to implement, cost effective and physically small. Hence the simple reflector antennas that are widely used today may no longer suffice in the future. A solution for this problem lies within the printed circuit technology, for its improved performance and cost effective nature. This thesis involved determining the most suitable printed circuit board antenna to use, designing it and testing it. [2, 3] Printed Circuit Board Antennas PCB antennas dates back to the early 1950 s, but due to the lack of good low cost microwave substrates the developments of these antennas were limited over the next 15 years. A rapid development in PCB antennas started in the 1970 s as thin, low cost antennas were needed for missiles and spacecrafts. PCB or microstrip antennas are an extension of the microstrip transmission line. A basic microstrip structure contains a thin substrate between 2 metal sheets which are usually copper. Microstrip line is then formed by etching away the necessary metal on top surface in so forming a strip. As long as the physical dimension of the strip and the relative dielectric constant remains unchanged, virtually no radiation will occur. By shaping the microstrip line into a discontinuity, power will radiate off from an abrupt end in the strip line. [2] Antenna Array A single antenna element is often limited to many applications because of its low gain and wide beam width. Where a higher gain, narrower beam width and increased range 2

16 are required, an array of antenna is to be implemented. In a physical terms, an antenna array would be multiple elements of antennas, in a linked configuration placed equal distances apart to achieve the desired higher gain. 1.2 Objectives The objectives of this thesis are: To determine the best suited PCB antenna for cellular devices. To propose a design and fabricate the chosen PCB antenna. To test and analyze the suitability of this antenna using the network analyzer and the power meter, thus measuring the impedance matching and the power radiated at 1.8GHz respectively. To design and fabricate an appropriate antenna array configuration for the antenna. To test and analyze the suitability of the antenna array by using the network analyzer and the power meter. To draw conclusions and make recommendations on how the PCB antenna built can be improved over the present design for the suitability of wireless communication devices at 1.8GHz. 1.3 Limitations of Research As an anechoic chamber was not available, measurements such as antenna gain, polarization, directivity and back radiations are effected by external objects and finite ground planes which cause reflection producing irregularities in radiation patterns. 3

17 This limits the accuracy of antenna measurement greatly. Far-field pattern measurements are not considered highly accurate in this thesis as antennas would need to be taken to a reflection free far-field testing range located in Gauteng, which is not available at UCT. 1.4 Source of Information The information and knowledge on which this report was based upon was acquired from Prof. B Downing, sites from the internet, Documents from the University of Cape Town as well as test and analysis that are conducted on the built antennas. 1.5 Plan of Development The report covers the following sections in the following order: Principles and concept involved in determining and constructing of a suitable PCB antenna. Proposed designs of the chosen antenna. Methods used in design and fabrication of antennas and antenna array. Test performed and error analysis on the built antennas and antenna array. Make conclusions and recommendation on the suitability of the antenna array and suggestions of improvements. 4

18 References [1] Sainati, Robert A. CAD of Microstrip Antennas for Wireless Applications. London: Artech House, Inc, 1996, pp. 1. [2] Chang, Kai. RF and Microwave Wireless Devices. Texas A&M University: John Wiley & Sons, Inc, 2000, pp. 1-3 [3] Professor B J. Downing 5

19 Chapter 2 Literature Review This chapter goes through the theories and properties of antenna, microstrip, balun and antenna array. These important factors were needed to be acknowledged, understood and practiced for this thesis. 2.1 Antenna Properties This section describes the important properties of antennas, all of which crucial when determining and designing antennas for applications. Since this is a practical project, there was no attempt to present more fundamental antenna theory Impedance The characteristic impedance of the transmission line that is connected to the antenna must equal input impedance of the antenna. Antenna resistance is also called radiation resistance. This factor is defined as the resistance that would dissipate as much power as the transmission line for which it is connected to. Antenna resistance is defined [1]: 6

20 P R = (2.1) 2 I Where R is the antenna resistance P is the power dissipated I is the current drawn from the antenna Wavelength The frequency at which the antenna operates at is dependant upon its wavelength. The following equation describes this relationship [1]: c λ = (2.2) f Where λ is the wavelength 8 c is the speed of light ( 3 10 m/s) f is the operating frequency of the antenna Gain This antenna property is used to compare differences in antenna radiation characteristics, both directivity and efficiency needs to be taken into account when determining antenna gain. This comparison is done by using a reference antenna of known gain, generally either a monopole antenna or a dipole antenna. Therefore when we talk about a gain of an antenna, it is meant the gain for which the antenna improves beyond the reference antenna. The gain of an antenna can be expressed as a 7

21 power ratio [2]: P ( ) 2 A db = 10log 10 (2.3) P1 Where A is the gain P 2 is the power of the reference antenna P 1 is the power of the actual antenna The power received by an antenna through free space can be expressed as: P r 2 PG t tgrλ = (2.4) π d Where P r is the power received P t is the power transmitted G t is the transmitting antenna gain G r is the receiving antenna gain λ is the Wavelength in meters d is the distance between the antennas, in metres Effective Area Effective area of an antenna can be said to be the measure of the effective transmitting or receiving area presented by an antenna. It is normally proportional, but less than, the physical area of the antenna. Effective are can be expressed as a relationship to gain by [1]: 8

22 2 λ A e = G (2.5) 4π Where A e is the effective area G is the gain of the antenna λ is the wavelength in meters Directivity Directivity is the term used when measuring how focused an antenna radiation pattern is. Directivity can be divided into two classes, Omni-directional and directional. Omni-directional antennas radiate and receive signals from all directions at the same time with the trade off of having a limited gain. Where directional antennas radiate and receive signals at a beam width that directed outwards from the antenna, either at one or opposite directions with a higher gain than omni-directional antennas. For any antenna, the directivity can be related to its effective area [2]: π e 2 4 A D = (2.6) 2 λ Where D is the directivity, and A e is the effective area λ is the wavelength in meters For small Beam width radiation, directivity can be also approximated by [3]: 9

23 27000 D = α α (2.7) 0.5horizontal 0.5vertical Where α 0.5horizontal is the horizontal half-power beamwidth α 0.5vertical is the vertical half-power beamwidth Efficiency This property is a measure of how much electrical power supplied to an antenna is transformed into electromagnet and due to losses (imperfect dielectrics, eddy current etc), not all energy transmitted to the antenna is radiated, with this the antenna efficiency is defined as [2]: P P transmitted r η = = (2.8) input R R + R r l Where η is the antenna efficiency P transmitted is the transmitted power P input is the input power R r is the antenna resistance R l is the resistance due to losses Polarization The electromagnetic wave that is radiated from an antenna is comprised of electric and magnetic field. These fields are orthogonal to each other and to the direction of propagation. The polarization of an antenna is described by the electric field propagated. In general, all electromagnetic fields are elliptically polarized as shown 10

24 from the figure below [2, 4]: Figure 2.1: Elliptically Polarized Electromagnetic Field Where E x is the electric field in the x-direction E y is the electric field in the y-direction E is the total electric field, sum of E x and E y. 2.2 Printed Circuit Board Antenna As mentioned before, PCB antennas are relatively simple structures where primary complicating factor are the relations between microstrip lines, dielectric substrates and the conducting ground plane. Although best accuracy can be achieved through rigorous mathematical and computational approaches, for this thesis, simpler and most widely used engineering methods were used. This section will describe all the microstrip technology and techniques applied [5]. 11

25 2.2.1 Types of Microstrip Antenna Three main types of PCB antennas were considered. Mainly the patch, dipole, and meander line antenna (MLA), the following table shows the characteristics of these antennas at 1.8GHz [5]: PCB Antennas Efficiency Directivity Physical Dimensions Patch High Directional Compact Dipole Medium Omnidirectional Compact Meander line Medium Directional Large Table 2.1: PCB Antenna Characteristics Microstrip Line The following figure shows the physical properties of a microstrip line: Figure 2.2: Microstrip Line The electric field lines are perpendicular to the microstrip line to the ground plane. 12

26 Most of the lines are concentrated below the strip where some lines partially extend into the air space above. These lines are parallel to the conducting surfaces of the microstrip lines. The magnetic field circles the microstrip line and doing so also extends into the air space above the strip. Both fields that exist in the air space above the strip reduce the effective dielectric constant seen by waves propagating along the line. If the fields above the strip did not exist, then the dielectric constant would the same as the substrate. The width and height of the microstrip line controls how much the dielectric constant depends on the substrate, this in term affects crucial microstrip parameters such as characteristics impedance and phase velocity, making them frequency dependent. These factors provides us with analysis techniques in calculating and designing microstrip lines, and hence PCB antennas. The following synthesis formula helps approximate the width of microstrip lines [6]: A = ( ε + 1) r (2.9) 4 ln 1 ε r 1 π B = + ln π 2 ε r ε r (2.10) 2 4h 4 C = ln (2.11) w h 59.95π D = (2.12) ε r A, B, C, D parameter defined for convenience in applying the design formula 13

27 ( B) Z 0 = A C (2.13) w h 2 πd 2πD ε r 1 πd = 1 ln 1 + ln π Z 0 Z 0 πε r Z ε r (2.14) Where w is the desired width of the microstrip line h is the height of the microstrip line above the ground plane Z 0 is the characteristic impedance ε r is the relative dielectric constant Due to difference in impedance with changing line width, impedance matching is between two different impedance elements is necessary. This can be done by placing a matching strip between the elements to be matched. The impedance of this matching strip can be expressed by: Z = (2.15) 0 Z1Z 2 Where Z 0 is the matched impedance Discontinuity Discontinuity is a term for a geometric change in the microstrip line referring to a gap, notch or a bend in the strip. This physical change alters the electric and magnetic field distributions of the strip resulting in energy storage and often radiation at the discontinuity. Due to these factors showing similar characteristics as elements such as capacitors and inductors, discontinuities can be expressed as the following equivalent circuits: 14

28 Figure 2.2.3: Discontinuity Equivalent Circuit 2.3 Balun The word balun is a contraction for balanced to unbalanced, where balanced means that both terminals of the feed must be of the same voltage level with respect to ground, if not then it is said to be unbalanced. This device is designed for systems where a centre-fed antenna (dipole for example) which is a balanced device, is fed to a coaxial line, which is unbalanced. The balance of the system is then upset as one side of the antenna is connected to the outer shield while the other is connected to the inner conductor. On the side connected to the shield, a current can flow across over the outer of the coaxial line. The fields that exist outside cannot be cancelled by the 15

29 fields from the inner conductor as the fields within can not escape, and hence the current flowing outside of the coaxial line will cause unwanted radiation. This is called line radiation. Baluns manipulate the voltages at the antenna terminals equal in amplitude with respect ground but opposite in phase, these voltages causes equal amount of current to flow on the outside of the coaxial line. Since the current are equal and out of phase with each other, they cancel out and hence reducing the effect of line current [7]. 2.4 Antenna Array In some antenna applications, higher gain and narrower beam width is required to increase range and to reject interference. Antennas can be arrayed to produce these characters. In this section, array theory and architecture is discussed Array Theory For derivation purposes, consider two identical antennas positioned on the x-axis in figure 2.4. The antennas are separated by a distance d. If the two antennas were in phase and applied with equal amplitude, Maxwell s equation states that the total field radiated by the antennas at distant points r (as distance at both distant points equal, denoting both distance vector r) is given by the sum of the fields from each antenna. Total field is expressed below [8]: E jkr jkr e e 1 (2.16) r r ( ) = E ( θ ) + E ( θ ) θ 2 Where E 1 and E 2 are fields radiated by the antennas 16

30 As the phase changes by 2π for every wavelength difference, the typical path length differences of the antennas are on order of a half-wavelength can be approximated by: ( θ ) d' d sin (2.17) The total field becomes: jkr e E 1 1+ r [ ] jkd sin( θ ) ( θ ) = E ( θ ) e (2.18) Figure 2.4.1: Array Array Architecture The simplest and easiest antenna array feed network to implement and design is the corporate fed array. With this type of configuration, all the elements in the array are connected in parallel linked by parallel power splitters. The amplitude levels of the antenna elements are controlled by way the power splitters are configured. The following figure shown is a typical four elements corporate fed array: 17

31 Power splitters Figure 2.4.2: Four Elements Corporate Fed Array 2.5 Design Consideration As my aim in this thesis is to determine and design a suitable antenna for the mobile cellular technology, the following important factors are considered: Antenna designed needs to be omnidirectional for effective reception on a mobile device. Array design and configuration needs to have the potential for simple implementation and installation. The antenna needs to be compact in physical dimension for compatibility in the mobile cellular industry. Cost effectiveness of the final product should be considered. 18

32 Reference [1] Antenna Terminology, 5 August [2] Antenna Definition, 5 August [3] Prof. M. Reineck, university of Cape Town. Lecture notes from course EEE497C, RF Components and Satellite Systems. [4] Antenna Polarization, 12 August [5] Chang, Kai. RF and Microwave Wireless Devices. Texas A&M University: John Wiley & Sons, Inc, 2000, pp [6] Roddy, D. Microwave Technology. New Jersey: Prentice Hall, Inc, 1986, pp [7] Dean Straw, R. The ARRL Antenna Book. Newington: The American Radio Relay League, 1994, Chapter [8] Sainati, Robert A. CAD of Microstrip Antennas for Wireless Applications. London: Artech House, Inc, 1996, pp

33 Chapter 3 Proposed Design As the antenna design needed to perform the requirements of the design consideration mentioned in the last chapter, I chose to design PCB dipole antennas due to its appropriate characteristics, these been omnidirectional, compact, cost effective. Mentioned in the last chapter, a dipole antenna need a balun to be compatible for connection to a co-axial cable, hence balun designs were also proposed. A Corporate fed array was chosen for my array design, as its simple configuration meant that manipulation of the arrays were possible if further testing were to be conducted for optimum antenna design. 20

34 3.1 Proposed PCB Dipole Radiator Design The following radiator design was proposed: Position x ¼ λ Figure 3.1: Proposed Dipole Radiator Design This half-wave PCB dipole design, as to all half-wave dipole, has two quarter-wavelength radiating elements. The radiating elements are connected to a ground patch by two quarter-wavelength strips, these strips provide the ground plane for the balun element on the other side of the substrate. This radiator design is the ground sheet of the whole dipole antenna, where it is connected to the ground/outer conductors of the co-axial cable. The copper sheet on the other side of the substrate will be the balun, where it is connected to the feed/inner conductor of the coaxial cable. These two sheets including the substrate in the middle will form the whole PCB dipole antenna design. [2] 21

35 3.2 Proposed Balun Design There are two balun designs implemented. One is the short circuit design and the other is the physically connected design. The following illustration shows the two designs: Chamfered bends ¼ λ ¼ λ Position z ¼ λ Position y ¼ λ a) Short circuit b) Physically connected Figure 3.2: Proposed Balun Designs In design a, the quarter-wavelength strips leading to the ground patch on the radiating sheet mentioned in 3.1 acts as a ground plane for the balun microstrip line. This microstrip line effectively has a gap in the ground plane between the two radiating element, this gap is terminated by an open circuit a quarter wavelength away at positioin y. The combination of these elements provides the balanced to unbalanced transition and the necessary impedance needed [2]. In design b, position z is physically connected to position x. This design is fundamentally similar to design a, except in this case it does not need to provide an open circuit quarter-wavelength away to terminate the gap, it is already physically 22

36 terminated [3]. The 45 o diagonal chamfered (or metered) bends on the microstrip lines shown in the figure above greatly minimizes the discontinuity reactance that would normally occur on a bend, as these chamfered bend are electrically shorter than the physical distance of the bend path. In an even more understanding perspective, a chamfered bend seems though if it is guiding the signal travelling through the path [1]. 3.3 PCB Dipole Antenna The figure below explains illustratively how the previously mentioned designs in this chapter are associated: Physical connection between two sheets Design a Design b Figure 3.3: Balun Designs 3.4 Array Design Considering that compactness in physical size is of importance, hence a suitable four 23

37 elements array was proposed. As in all corporate arrays, the antenna elements are connected in parallel with linked parallel power splitters, hence the matching of parallel impedance additions and antenna impedances are considered. The distances between each element are, as supervised, is to be half a wavelength apart. Further test can be conducted by altering the distance between array elements for optimum antenna performance. The design below illustrates how the impedance matching, array distances and chamfered bends are accounted for: λ/2 λ/2 Antenna Elements Impedance matching Strips Chamfered bends Figure 3.4: Proposed Array Design 24

38 References [1] Sainati, Robert A. CAD of Microstrip Antennas for Wireless Applications. London: Artech House, Inc, 1996, pp [2] SPAS Hard Ware Design Antenna Array, 21 July S/antenna.html [3] 2.4GHz Diversity Polarization Diversity Antenna, 13 August ole_antenna.pdf 25

39 Chapter 4 Design and Fabrication Methods This chapter describes the design procedures taken in assisting linking knowledge between the theoretical understandings, and the practical fabrication of antennas. 4.1 Design Procedures The following design procedures were implemented, taking into consideration the time and cost constraints: Design and building of prototype microstrip dipole antennas. Analyzing the performance of prototype antennas in helping designing the etched PCB dipole antennas. Design and fabrication of PCB dipole antennas. Design and fabrication of antenna array panel. 26

40 4.2 Building and Testing of Prototype Microstrip Antennas The reason for building these initial antennas are so that the relationship between antenna theory and practical antenna building could be more understood before going ahead in designing the final PCB antennas. Further, it is possible to tune these antennas by trimming the copper strip conductors and adjusting the dimensions. Three prototype antennas were built using different dimensions and methods, this assists in determining which type would be more feasible for further design. All three antennas were fabricated using 0.5mm copper plate as conductive sheet and 6mm polypropylene substrate Prototype Designs The width the microstrip lines used in these antennas are calculated using the microstrip line synthesis equation (4.16), with the following known dimensions: Operating Frequency: 1.8GHz Substrate relative permittivity: 1.42 Height: Impedance a) 6mm 61Ω b) 75Ω Calculated width = a) 15.08mm b) 20.6mm The length of the dipole antennas are λ/2, which is 83mm. This is split into two radiating elements λ/4 in length (41.5mm), separated by a 5mm gap. 27

41 Design A: 75Ω 61Ω Figure 4.1.1: Prototype Design A In design A, balun width is equal to the width of its ground plane, forming a 1:1 ratio between balun to ground [2]. Design B: 75Ω 61Ω Figure 4.1.2: Prototype Design B This design used a balun to ground plane width ratio of 3:1 as advised in the microstrip CAD, it is said that dipole antenna of such design need to implement this ratio to operate properly [1, pp ]. Ratio 3:1 Design C: Figure 4.1.3: Balun to Ground Plane Width Ratio 3:1 28

42 75Ω 61Ω Figure 4.1.4: Prototype Design C This design implemented the physically connected balun, using balun to ground plane width ratio of 1:1 [3] Prototype Performance The antenna performances below were taken from the 2 S 11 readings on the network analyzer. Design A db MHz Figure 4.1.5: S11 2 of Prototype Design A 29

43 Design B db MHz Figure 4.1.6: S11 2 of Prototype Design B Design C db MHz Figure 4.1.7: S11 2 of Prototype Design C 30

44 Prototype Centre Operating Band Minimum Maximum Frequency Power Power Reflected Reflected Design A 1719 MHz MHz 20 db 40 db Design B 1280 MHz MHz 15 db 18 db Design C 2180 MHz MHz 7 db 22 db Table 4.1: Prototype S11 2 Analysis From the result shown above, we can see that neither the 1:3 balun to ground width ratio nor the physically connected balun had the satisfactory result, hence the designs for etched PCB antennas will be implemented using design A: Quarter-wavelength short circuit balun with 1:1 balun to ground width ratio. 4.3 Designs of PCB Antennas The designs of the PCB antennas in this section are developed from design A of the prototype antennas, which proven to be the best balun configuration. At this stage, the aim is to design the best suited antenna with respect to its bandwidth. For a dipole antenna, the bandwidth is mainly dependent upon the width of the radiating elements, hence four PCB dipole antennas with different dimensions (8mm, 10mm, 12mm, wideband) has been designed to understand this relationship further, and from this, the best suited antenna were chosen for the array elements [4]. 31

45 Microstrip line widths are calculated by using the synthesis equation (2.16) Operating Frequency: 1.8GHz Substrate relative permittivity: 4.8 Height: Impedance a) 1mm 61Ω b) 75Ω Calculated width = a) 0.18mm b) 0.8mm The following antennas were drafted using AutoCad software, laser printed out in scale 1:1 format and given to an etching service for it to be printed onto carbon films. They then etched the designs onto printed circuit boards. 8mm Dipole: 8mm 61Ω 75Ω Figure 4.2.1: 8mm PCB Dipole 32

46 10mm Dipole: 10mm 61Ω 75Ω Figure 4.2.2: 10mm PCB Dipole 12mm Dipole: 61Ω 12mm 75Ω Figure 4.2.3: 12mm PCB Dipole Wideband Dipole: 61Ω Wideband 75Ω Figure 4.2.4: Wideband PCB Dipole 33

47 4.4 Array Design As mentioned in the last chapter, the impedance matching and distances between elements and power splitters are of crucial importance. By using the microstrip line synthesis equation, the following widths are calculated: Operating Frequency: 1.8GHz Substrate relative permittivity: 4.8 Height: Impedance a) 1mm 50Ω b) 100Ω c) 70.1Ω (matching of 50Ω to 100Ω) Calculated width = a) 2mm b) 0.4mm c) 0.9mm X X X 50Ω 70.1Ω 100Ω Figure 4.3.1: Corporate Fed Array Power Splitter 34

48 At points X, the two parallel 100Ω line joins giving an effective impedance of 50Ω. These points are then matched to other 100Ω s, resulting the 50Ω needed at the feed point of the array. The lengths of the parallel 100Ω lines are all equal to provide equivalent phase to all signals passing through. As mentioned previously, the shown chamfered bends and joints have been implemented to minimize discontinuity reactance [1]. The array power splitter was also drafted in AutoCad software and fabricated the same way as the PCB antennas. 35

49 References [1] Sainati, Robert A. CAD of Microstrip Antennas for Wireless Applications. London: Artech House, Inc, 1996, pp 28. [2] SPAS Hard Ware Design Antenna Array, 21 July S/antenna.html [3] 2.4GHz Diversity Polarization Diversity Antenna, 13 August ole_antenna.pdf [4] Prof. M. Reineck, university of Cape Town. Lecture notes from course EEE497C, RF Components and Satellite Systems. 36

50 Chapter 5 Tests and Analysis of PCB Dipole Antennas and Antenna Array Tests and analysis was carried out on the etched antennas and antenna array after it was fabricated. The tests conducted include Impedance matching of all etched products, and the power radiation of the antenna array and the antenna employed. 5.1 Impedance Matching of Etched Antennas The S11 2 reading shown on the network analyzer determines how well the impedance is matched on the component that it is connected to. The test is an indicator of the reflected power which analyzes whether the component is radiating at the correct frequency, and its radiating bandwidth. 37

51 5.1.1 S11 2 Test on PCB Dipole Antennas The S11 2 test was conducted on the following antennas: 8mm Dipole: S11 2 of 8mm Dipole db MHz Figure 5.1.1: S11 2 of 8mm PCB Dipole 10mm Dipole: S11 2 of 10mm Dipole db MHz Figure 5.1.2: S11 2 of 10mm PCB Dipole 38

52 12mm Dipole: S11 2 of 12mm PCB Dipole db MHz Figure 5.1.3: S11 2 of 12mm PCB Dipole Wideband Dipole: S11 2 of Wideband PCB Dipole db MHz Figure 5.1.4: S11 2 of Wideband PCB Dipole 39

53 PCB Dipole Antenna Center Frequency Operating Band Minimum Power Reflected Maximum Power Reflected 8mm 1355MHz MHz 8dB 22dB 10mm 1385MHz MHz 18dB 60dB 12mm 1400MHz MHz 10dB 18dB Wideband 1370MHz MHz 5dB 14dB Table 5.1: PCB S11 2 Analysis Error Analysis Although all the antennas showed the expected bandwidth characteristics, it is not operating at the ideal frequency of 1.8GHz. This error can arise from the possible inaccurate information on the value of the dielectric permittivity, inexact etching, and the reflected loss at the connectors. From the readings taken, it is clear that the 10mm PCB dipole is the best impedance matched antenna design fabricated, and the fact that in this thesis wide bandwidth is not a requisite, the 10mm dipole were the antennas chosen for implementation of array elements Frequency Correction on the 10mm Design We know that the operating frequency of the antenna is dependent on the length of the radiating element and the length of the quarter-wave matching strip by the equation c L =. Hence by, trial and error, decreasing the length of these two components I 4 f 40

54 was able to increase the operating frequency to the desired value of close to 1.8GHz. The frequency correction is shown below: Decreased Length Figure 5.1.5: 10mm PCB Dipole after Frequency Correction 5.2 Antenna Array Test and Analysis In this section the performance of the array element, array power splitter and the antenna array will be discussed and analyzed. 41

55 5.2.1 Array Element Performance As the array is using the corrected 10mm antenna for its radiating elements, both the reflected power S11 2 and the polar radiation of this antenna will be tested. The S11 2 test result of the 10mm antenna after frequency correction is shown below: S11 2 of 1.8GHz 10mm Dipole db MHz Figure 5.2.1: S11 2 of 10mm PCB Dipole after Frequency Correction In the figure above, although the reflected power is lower than the original 60dB, the center frequency has been corrected to 1.74GHz, which is close enough for 1.8GHz applications. As an anechoic chamber was not available for use in the polar radiation testing, the test took place on the roof of the engineering building to minimize reflection. The following illustration will show how reflections have being minimized by the placing equipments in a diagonal configuration. 42

56 Antenna under test Open space Figure5.2.2: Polar Radiation Test Configuration Theoretically by using the above configuration, all power received will not have any reflected power. However, interference was detected which affected the received signals. These signals were most likely from Cell C (1.8GHz), MTN (900MHz), Vodacom (900MHz) and the SABC (UHF and VHS). The polar radiation test result shown indicates that the antenna designed is omnidirectional and due to the ground plane located behind the radiating element, it has a lot less back radiation as expected: Figure 5.2.3: Polar Radiation of an Array 180 Polar Radiation of 10mm 1.8GHz Dipole element 270 Power radiated in µw 43

57 5.2.2 Array Power Splitter Before connecting the four array elements on to the power splitter, the power splitter needed to undergo the S11 2 test to give an indication on how well it is matched. Impedance terminators were connected to prevent power radiation at each splitter ends. The terminator connections and the S11 2 result are shown below: Terminators Figure 5.2.4: Impedance Terminator Connections S11 2 of Array Power Splitter db MHz Figure 5.2.5: S11 2 of Array Power Splitter 44

58 The result indicates that the power splitter is matched closely to the 1.8GHz region, but not exact due to factors such as loss in connections and imprecise etching Antenna Array Performance The array elements were connected to the power splitters by four identical length cables (40cm) to ensure no phase change. Each array elements were positioned half wavelength apart, as this will have enough distance so that the coupling between the elements is minimal. Antenna array components are then attached on to a non-reflective material called polycarbonate to ensure a minimal reflective structure. The figure below will show how the antenna array physically stands. Figure 5.2.6: Antenna Array 45

59 The S11 2 and the polar radiation test are illustrated below: S11 2 of 1.8GHz Antenna Array MHz db Figure 5.2.7: S11 2 of Antenna Array Polar Radiation of Dipole Array Power radiated in µw Power radiated in µw Figure 5.2.8: Polar Radiation of Antenna Array 46

60 From the S11 2 test we see that although the impedance of the antenna array is matched at 1.8GHz, it is better matched at a higher frequency. This is caused by the mismatches between the array elements, and the mismatch previously determined in the power splitters. The polar radiation shows us that the antenna array has retained its omnidirectivity as expected, with gain in the forward and side radiation. The back radiation in the antenna array has also been limited; this is due to ground plane reflections from the array power splitter. Theoretically, the antenna array will have the most gain in the vertical region. 47

61 Chapter 6 Conclusions and Recommendations This chapter discusses the conclusions that have been drawn from the results of tests conducted on the antenna array. Recommendations are then made with regard to the conclusions. 6.1 Conclusions The results from the tested antenna array are satisfactory and have achieved the objective set for this thesis, which is to design and build a high gain omnidirectional printed circuit board dipole antenna array operating at 1.8GHz Results of Array Elements The reflected power test (figure 5.2.1) indicates that the tested array element is closely matched to 1.8GHz at 1.74GHz, with an excellent reflected power of 60dB. 48

62 The polar diagram of the power radiation (figure 5.2.3) shows that the tested array element is radiating as a dipole antenna (omnidirectional), with the exception of minimal back radiation. This is caused by reflections from the ground plane connected to the antenna radiating elements Results of Array Power Splitter The reflected power test (figure 5.2.5) indicates that array power splitter is well matched at 1.88GHz. Although close to the 1.8GHz, but poorly matched at 1.8 GHz exactly. This is caused by poor connections between the power splitters and the cables. Another reason for this mismatch can be the result of imprecise etching of the microstrip lines Results of Antenna Array The reflected power test (figure 5.2.7) indicates that the antenna array is well matched at 1.8GHz, but better matched at 1.95GHz. This is caused by mismatches between each array elements and the poor matching of the power splitter. The polar radiation test (figure 5.2.8) shows that the antenna array has retained its omnidirectivity with the advantages of increased gain in the forward and side radiation. Back radiation is limited due to the reflections discussed in section 6.1.1, and the reflections from the power splitter. From the array configuration implemented, the maximum gain theoretically occurs in the vertical region, where the polar diagram is a test conducted for power radiation at the azimuth plane. 49

63 6.2 Recommendations The following recommendations on the improvements, testing and possible application of the antenna array have been made from the drawn conclusions. Further testing on the current antenna array has also been proposed Improvements on the Antenna Array The following recommendations are made to improve the performance of the antenna array: Instead of printing designs on to scale 1:1 photo copies, rather give the etching service a software form of the design. These software designs can be done in either using Gerber or Protel software which are capable of designing printed circuit boards. This will greatly improve the precision of etching, hence reducing the impedance mismatches that will occur. The connections used between each array elements and the power splitter are BNC connectors which were not designed specifically for 1mm PCB boards, hence the heads needed to be physically altered to fit the boards. Therefore for future designs, reflective loss can be greatly minimized if connectors of the correct dimensions were used. It has also been proven that SMA connectors are better suited for microwave applications than BNC connectors. 50

64 Back radiation can be increased for optimum omnidirectivity by altering the position of the power splitter. Placing it below the array elements so that reflections can be minimized on the azimuth plane Antenna Testing Recommendations The following should be implemented to improve accuracy of the test results: An anechoic chamber will greatly improve the accuracy of radiation tests conducted on antennas, as reflections are reduced. An automated stepping turn table would help with taking polar radiation readings, as each reading taken will be stepped at an accurate angle Possible Antenna Array Application As the antenna array built is omnidirectional with high gain in the vertical region, it could be implemented in the case where ground to air communications were needed at 1.8GHz Proposed Further Testing The following tests and design alteration on the antenna array are proposed: The distance between the array elements can be altered to test for optimum array performance. 51

65 The configuration on how the array elements can be changed for different applications, where increased antenna gain in one direction can be a trade off for antenna omnidirectivity. The physical size can be reduced for commercial use. This can be done by replacing materials used, and possible design alteration. 52

66 Bibliography Sainati, Robert A. CAD of Microstrip Antennas for Wireless Applications. London: Artech House, Inc, 1996 Chang, Kai. RF and Microwave Wireless Devices. Texas A&M University: John Wiley & Sons, Inc, Roddy, D. Microwave Technology. New Jersey: Prentice Hall, Inc, Dean Straw, R. The ARRL Antenna Book. Newington: The American Radio Relay League, Prof. M. Reineck, university of Cape Town. Lecture notes from course EEE497C, RF Components and Satellite Systems. 53

67 Appendix A Substrate Data Sheet The data sheet shown in this section was the information given by Northtech etching service to help determine the characteristics of the substrate. 54

68 55

69 Appendix B Photographs of S11 2 Scope Displays The S11 2 test results were taken on a digital camera, then converted the results into Microsoft excel format. The original photographs are shown in this section. 56

70 Figure B1: Photograph of Prototype Design A S11 2 Test Result Figure B2: Photograph of Prototype Design B S11 2 Test Result 57

71 Figure B3: Photograph of Prototype Design C S11 2 Test Result Figure B4: Photograph of 8mm PCB Dipole S11 2 Test Result 58

72 Figure B5: Photograph of 10mm PCB Dipole S11 2 Test Result Figure B6: Photograph of 12mm PCB Dipole S11 2 Test Result 59

73 Figure B7: Photograph of Wideband PCB Dipole S11 2 Test Result Figure B8: Photograph of 1.8GHz PCB Dipole S11 2 Test Result 60

74 Figure B9: Photograph of Array Power Splitter S11 2 Test Result Figure B10: Photograph of Antenna Array S11 2 Test Result 61

75 Appendix C Carbon Film Copies of Etched Printed Circuit Boards Antenna designs printed on carbon films were used in the etching process. These films are shown in this section. 62

76 63

77 64

78 65

79 66