ELE-6216 ANTENNA PROJECT

MICROSTRIP PATCH ANTENNA DESIGN

Submitted to: Jouko Heikkinen Department of Electronics 04. April 2011

i

ABSTRACT TAMPERE UNIVERSITY OF TECHNOLOGY Master’s Degree Programme in Radio Frequency Electronics SAINJU, PRABHAT MAN BANIYA, ANIL: Microstrip Patch Antenna Design Project Report, 21 pages, 3 Appendix pages April 2011 Subject: ELE-6216 Antenna Project Examiner: Jouko Heikkinen Keywords: Microstrip Patch Antenna, Antenna Design, Antenna Simulation Microstrip patch antennas are widely used in modern day communication devices due to

their small size, low cost and ease of fabrication. This report deals with the design, con-

struction, simulation and measurement of a microstrip patch antenna. The antenna has

been constructed to operate under given design specifications. A standard commercial

tool has been used to simulate the antenna structure. The simulations have been com-

pared with the real time measurement data. The data have been properly analysed to

justify the theory.

ii

TABLE OF CONTENTS

Abstract .............................................................................................................................. i

Terms and Abbreviations ................................................................................................. iii

1 Introduction ............................................................................................................... 1

2 Theory of Microstrip Antennas ................................................................................. 2

2.1 Antenna Parameters .......................................................................................... 4

2.1.1 Antenna Gain ....................................................................................... 4

2.1.2 Return Loss .......................................................................................... 4

2.1.3 Antenna Efficiency .............................................................................. 4

2.1.4 Antenna Bandwidth ............................................................................. 4

3 Design and Simulation .............................................................................................. 5

3.1 Design ............................................................................................................... 5

3.1.1 Design Specification ............................................................................ 5

3.1.2 Substrate Specification ........................................................................ 5

3.1.3 Feed Specification................................................................................ 5

3.1.4 Calculations ......................................................................................... 5

3.2 Simulation Tool and Methodology ................................................................... 6

3.2.1 Simulation Components ....................................................................... 6

3.2.2 Simulation Results ............................................................................... 6

4 Fabrication................................................................................................................. 8

5 Measurements ........................................................................................................... 9

5.1 VNA Measurements .......................................................................................... 9

5.2 Measurement with SATIMO Starlab .............................................................. 10

6 Analysis of AUT ..................................................................................................... 12

7 Conclusion .............................................................................................................. 13

References ....................................................................................................................... 14

Appendix A: Simulated Results ...................................................................................... 15

Appendix B: Measured Results of AUT ......................................................................... 16

iii

TERMS AND ABBREVIATIONS

ADS Advanced Design System

MPA Microstrip Patch Antenna

RF Radio Frequency

EMC ElectroMagnetic Compatibility

TUT Tampere University of Technology

VNA Vector Network Analyzer

FEM Finite Element Method

AUT Antenna Under Test

1

1 INTRODUCTION

Microstrip Patch Antenna is a flat radiating element mounted on a flat surface. The ra-

diating element consists of a rectangular metallic patch placed over larger metallic sheet

that acts as a ground plane. The two sheets are separated by a dielectric. This structure

of a dielectric slab placed between the two metallic sheets forms a microstrip line. Mi-

crostrip patch antennas are easy to fabricate via photo-lithographic process or printed

electronics and the design concerns are merely limited to the scaling of the dimensions.

These antennas are highly miniaturized and mostly advantageous on modern day wire-

less communications where size does matter.

2

2 THEORY OF MICROSTRIP ANTENNAS

Microstriplines are planar transmission lines with a conducting metallic strips realized

on flat surfaces over a ground plane separated by a dielectric substance. These micro-

strip structures can easily be constructed using the printed electronics and photolitho-

graphy technology. This microstrip structure when properly scaled in dimension and

appropriately fed acts as a radiating element. This feature has now been widely utilized

in modern day communication devices since they are miniaturized compared to other

antenna structures where space and size are also important factors to be taken care of.

They are low cost and easily fabricable.

An important part of the antenna apart from its own structure is the feed; its the source

of the signal that is required to be radiated to the outer space.

Types of microstrip antennas

Rectangular

Circular

Planar Inverted F Antenna (PIFA) etc.

Among the above listed types, this report focuses in the rectangular microstrip antennas.

Figure 3.1. MPA with probe feed

Figure 3.1 shows the rectangular patch antenna. Upper layer of the structure consist of

the radiating element with length (L) and width (W) which is perfect electric material. In

our case the patch is square hence L=W. The lower layer of the structure is also a per-

3

fect electric material known as ground plane and the middle layer is the substrate of

height h.

There are different options available for the feed sources. The most used sources are:

Inset Microstrip

Quarter-wave transformer

Probe coupling

The dimensions of the rectangular patch antennas affect the overall operations of the

antenna and are directly related to the antenna specifications.

With the center frequency of the antenna given, the width of the microstrip patch is tak-

en as [1]

𝑊 =𝐶0

2 × 𝑓0 𝜀𝑟 + 1

2

3.1

Where 𝜀𝑟 = 𝑑𝑖𝑒𝑙𝑒𝑐𝑡𝑟𝑖𝑐 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡 𝑜𝑓 𝑡ℎ𝑒 𝑠𝑢𝑏𝑠𝑡𝑟𝑎𝑡𝑒

C0 = velocity of light in free space

𝑓0 = 𝑐𝑒𝑛𝑡𝑒𝑟 𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦 𝑜𝑓 𝑎𝑛𝑡𝑒𝑛𝑛𝑎

The length extension of the rectangular patch is given as

∆𝑙 = 0.412 × ℎ × 𝜀𝑒𝑓𝑓 + 0.3 (

𝑊ℎ

+ 0.264)

𝜀𝑒𝑓𝑓 − 0.258 (𝑊ℎ

+ 0.8) (3.2)

Where the effective dielectic constant 𝜀𝑒𝑓𝑓 is given as,

𝜀𝑒𝑓𝑓 = 𝜀𝑟 + 1

2+ 𝜀𝑟 − 1

2 1 + 12

ℎ

𝑊 −1

2

3.3

And h is the height of the dielectric substrate.

Thus finally the length of the patch L is given as,

𝐿 =𝐶0

2𝑓0 𝜀𝑒𝑓𝑓− 2∆𝑙 (3.4)

As mentioned earlier, feed position is another important portion of the antenna design.

Of the lists of feeds mentioned, probe coupling has been focused in this report. The lo-

cation of the feed point from the edge of the patch is given as,

∆𝑥 =𝐿

𝜋cos−1

𝑍𝑖𝑛𝑍𝐴 3.5

Where Zin is the input impedance of the antenna and ZA is the antenna impedance given

by

4

𝑍𝐴 =90𝜀𝑟

2

𝜀𝑟 − 1 𝐿

𝑊

2

𝑂ℎ𝑚𝑠 3.6

2.1 Antenna Parameters

2.1.1 Antenna Gain

Antenna gain is defined as the ratio of electric field intensity of an antenna in a given

direction to that of an isotropic antenna that radiates equally in all directions.

2.1.2 Return Loss

The return loss refers to the logarithmic ratio of measured in dB that compares the pow-

er reflected to the antenna to the power fed to the antenna by the feed [3].

Return Loss

𝑅𝐿 = −20 log 𝜏 𝑖𝑛 𝑑𝐵𝑠 3.7

Where reflection coefficient

𝜏 =𝑍𝐴 − 𝑍0

𝑍𝐴 + 𝑍0 (3.8)

2.1.3 Antenna Efficiency

Antenna efficiency is defined as the ratio of radiated power to the supplied power or the

available power [4].

2.1.4 Antenna Bandwidth

Antenna bandwidth is the frequency range within which the antenna operates at its op-

timum level. Analytically it can be taken as the range of the frequencies where the re-

turn loss is > 10 dB.

5

3 DESIGN AND SIMULATION

3.1 Design

This section includes the design methodology implemented for the construction of the

microstrip patch antenna. Guided work instruction has been followed for the Design

specification [5].

3.1.1 Design Specification

Following are the specifications that the designed microstrip antenna has to meet:

Square Microstrip Patch

Center frequency = 2 GHz

Bandwidth > 1 %

Input return loss > 10 dB ( 50 ɏ )

Efficiency > 70 %

Gain > 4 dBi

3.1.2 Substrate Specification

The substrate to be used here is : Rogers RO4003 (with photoresist coating) , thickness

60 mils (1.524 mm) [6].

3.1.3 Feed Specification

The feed used in this case is probe coupling. The probe coupler used here is 50 Ohm

SMA 4-Hole Flange Mount Jack Receptacle - Flush Dielectric [7].

3.1.4 Calculations

Width of the patch W (as per equation 3.1) = 50.60 mm

Length of the patch L (as per equation 3.2, 3.3 and 3.4) = 40.39 mm

Probe feed location (∆𝑥) = 15.73 mm

Since the required microstrip should have square orientation, the length (L=W) has been

taken as the dimension of the patch.

The required spacing of the substrate and ground plane around the patch is L/4. Hence

the length of the square substrate and the square ground plane is 60.58 mm. Please note

that this dimension of substrate and the ground plane is in reference to the patch dimen-

sion.

6

3.2 Simulation Tool and Methodology

Simulation tool used in this case is Agilent EMPro 2010.07 version. FEM Iterative Ma-

trix Solver was chosen as the numerical method with delta correction of 0.005 for high-

er accuracy.

3.2.1 Simulation Components

The components of the microstrip antenna that was simulated using EMPro were:

Ground Plane (taken as Perfect Electric Conductor)

Substrate (With Relative Permittivity 3.38 and loss tangent 0.0021)

Patch (radiating element assumed to be Perfect Electric Conductor)

Pin of the probe feed (Source port assumed to be Perfect Electric Conductor.

The source is fed from the center of the pin to the edge of the ground plane).

3.2.2 Simulation Results

Simulations were conducted based on hit and trial method as the calculated dimensions

did not exactly yield the required specifications. Manipulation was mainly done with the

length of the square patch and the probe feed location. Following are the few of the ma-

jor simulation results obtained from EMPro for different combinations:

Table 4.1. Effect of variation of dimension of patch on antenna parameters

Length

(L=W in

mm)

Δx

(in mm)

Center Fre-

quency f0 (in

GHz)

Bandwidth

(in MHz)

Return

Loss

(in dBa)

Gain

(in dBi)

Efficiency

(in %)

40.39 15.73 1.905 - 8 3.78 35

39 11 1.953 25.4 25.205 4.706 37

38.4 11 2.042 28.1 25.969 4.706 44

38.6 11 2.033 28 25.205 4.703 55

38.7 11 2.0104 27.5 22.280 4.694 83

As from table 4.1 it is clear that the closest dimension that matches the given specifica-

tions is the last record with length 38.7 mm and probe feed point placed at 11 mm from

the end of the patch. The center frequency observed at this point was at 2.010 GHz with

bandwidth 27.5 MHz. Return loss 22.28 dBa and gain of 4.694 dBi. Hence this dimen-

sion has been used for the analysis and construction of the microstrip antenna.

The scattering parameter s11 gives the reflection coefficient of the antenna. Figure 4.1

shows the variation of reflection coefficient against the frequency for the dimension

selected above.

7

Figure 4.2. Reflection coefficient versus Frequency

Figure 4.2 shows the s11 parameter variation with the frequency for the above selected

dimension. The center frequency has been marked at 2.0104 GHz where the measured

return loss was 22.28 dBa. Similarly, the bandwidth was measured at the points where

the return loss was 10 dB. These points have been marked in the figure above.

The smith chart presented in Figure A.1 on appendix A shows the simulated resonant

frequency at 2.0078 GHz where the impedance was found to be (0.10607-0.00011937i)

ohms.

Similarly the Figure A.2 shows the radiation pattern of EM field in three dimensions.

The major lobe is protruding above the patch while a small back lobe is also visible be-

low the ground plane.

8

4 FABRICATION

The antenna dimension once fixed was built on the substrate Rogers RO4003 by wet-

etching process. The photoresist mechanism used was positive photoresist. Following

steps were followed for the fabrication process:

The geometry of the microstrip patch was exported from EMPro to Agilent

ADS. Then geometry was printed on the standard A4 transparent sheet. Also the

geometry of the ground plane was similarly printed on the other transparent

sheet.

The substrate was cut to the estimated dimension and its protective cover was

removed.

The substrate was placed between the two transparent sheets such that one side

of substrate faced the print of the patch and the other side faced the print of the

ground plane.

Then the enclosed sheet was placed inside the UV Exposure unit where the sub-

strate was exposed for 1 minute.

The exposed substrate was then placed on the bath of development chemical

NaOH. The exposed part should now become visible. This is the part that has to

be etched out.

The developed substrate was then placed in the bath of etching chemicals; water

(H2O) + hydrochloric acid (HCl) + hydrogen peroxide (H2O2).

The excess copper layer was removed from the radiating side with only the

patch left. An empty circle created for the probe feed was visible. The ground

plane however was intact with the circular hole for probe feed etched out.

Any remaining unnecessary copper on the substrate was mechanically removed.

A through hole was drilled at the circular mark created on both the patch and

ground plane. This would serve for the connection of the probe feed to the ra-

diating element.

The Flange Mount Jack was soldered with the pin of the jack soldered to the ra-

diating patch and the base soldered to the ground plane.

9

freq (1.800GHz to 2.200GHz)

Imp

ed

an

ce

Readout

m1

m1freq=s1_1=0.258 / 0.758impedance = 84.845 + j0.622

2.000GHz

5 MEASUREMENTS

5.1 VNA Measurements

The first measurement carried out was with the 8722D HP VNA. The effect of the cable

and connector of the VNA on impedance and the overall delay was considered during

the measurement. To reduce its effect, VNA was calibrated accordingly.

The result was obtained in the form of smith chart. It showed the variation of input im-

pedance of AUT with the frequency. The frequency range tested was from 1.8 GHz to

2.2 GHz. At 2 GHz (the expected center frequency of the AUT), the input impedance

observed was 84.845+j0.622 Ω. The center frequency was found approximately near to

the given specification of 2 GHz.

Figure 6.1. Smith Chart of Microstrip Antenna as observed from VNA

The return loss and bandwidth measured by VNA is shown in figure 6.2. Markers have

been placed at the point of interest. The center frequency was found to be at 2.004 GHz

and bandwidth was found to be 23 MHz (>1%) which matched our expected perfor-

10

mance. Return loss at the center frequency was found to be -12.06 dB which is better

than expected (<-10 dB).

Figure 6.2. s11 (Scattering Parameter) versus Frequency Plot as observed from VNA

It was also noted that the impedance of the antenna is greatly affected by the obstacles

in its near field region. The effect was not noticed when the obstacle was in the far field

region or beyond the main lobe of the antenna.

5.2 Measurement with SATIMO Starlab

It is a test station for the 3D measurements of the antenna radiation pattern. StarLab

uses an indirect measurement technique based on the Near Field measurement of the

AUT. It reconstructs the far field from the sampled near field using ”Near field to far

field transformation” algorithm [8].

1.85 1.90 1.95 2.00 2.05 2.10 2.151.80 2.20

-12

-10

-8

-6

-4

-2

-14

0

Frequency (GHz)

S11 (

dB

)

Readout

m1

Readout

m2

Readout

m3

m1freq=dB(s1_1)=-9.910

1.993GHzm2freq=dB(s1_1)=-9.781

2.016GHz

m3freq=dB(s1_1)=-12.069Min

2.004GHz

11

Figure 6.3. E and H Field gains at different zenith angle

Figure 6.3 shows the variation of the linear gain at different zenith angles (θ) for E and

H fields. The figure clearly depicts the main and back lobe of the AUT. The back lobe is

sufficiently small. This small back lobe radiation is an added advantage for using the

antenna in the cellular devices. [9]

Rest of the figures has been included in the appendix B.

Figure B.1 shows the efficiency of antenna at different frequencies. It is clear that AUT

has the highest efficiency at the center frequency.

Figure B.2 shows the variation of E and H field gains at different zenith angle in linear

form. It is similar to the figure 6.3 except that the figure 6.3 is the polar plot of the same

data.

Figure B.3 shows the gain of antenna at different frequencies. It is clear that AUT has

the highest gain at the center frequency.

12

6 ANALYSIS OF AUT

The entire antenna project included the design, simulation, fabrication and measurement

of the patch antenna. The design itself was not perfectly implemented in the fabrication.

The first simulation result in table 4.1 shows that the theoretically calculated dimensions

did not meet the required specifications.

The center frequency itself was far from the specified value. Different hit and trials

were performed with the variations in length and the probe feed locations. The most

approximate result was found with the simulation and was implemented in the fabrica-

tion. The fabricated antenna with the given dimension yielded a quite satisfying result

compared to the specifications. The table below shows the comparative analysis of theo-

retical, simulated and fabricated antenna model.

Table 7.1. Analysis of the AUT data

Specifications

Square Microstrip Patch

Center frequency = 2 GHz

Bandwidth > 1 %

Input return loss > 10 dB (50 Ω)

Efficiency > 70 %

Gain > 4 dBi

Design Parameters

Length =38.7 mm

Δx = 11 mm

Height of substrate = 1.524

mm

Parameter Value from Simulation

Model

Measured from AUT

Center Frequency (GHz) 2.0104 2.004

Bandwidth (MHz) 27.5 23

Return Loss (dB) 22.28 12.069

Efficiency (%) 83 77.6

Gain (dBi) 4.694 5.101

The simulation data seems to match with the required specification. Hence the simu-

lated design was approved for the fabrication. Similarly the AUT measurement data also

matched the required specifications. However, while comparing these two sets of data,

bandwidth, return loss and efficiency of measured data seem to be less than the simu-

lated data within the acceptable limits. The gain towards the main lobe however is

greater than the simulated one.

13

7 CONCLUSION

Thus, a microstrip patch antenna was designed, simulated and constructed with the wet

etching process. The antenna was then measured in the antenna measurement sites to

yield the standard parameter data. An important take away from this project was the

learning of standard simulation tools like Agilent EMPro and Agilent ADS. Involve-

ment of team members from the design till the measurement made the every aspect of

modern day antenna design clear.

The deviation of the design results as per Table 4.1 was not clearly understood though.

The dimensions calculated from the standard analytical formulas did not yield the exact

specifications and thus hit and trial method was performed varying the length and probe

feed location. However, no clear trend was found between dimension and the center

frequency of the antenna. The numbers of simulations performed were not enough to

conclude on any specific reason for the variation of data observed in the simulation.

Higher delta error accuracy level would have yielded more accurate result in the cost of

higher computation time and processing power.

The project gave an insight on how the modern day antennas are constructed and what

are the things that need to be taken into account for the construction of antenna under

given specification and how the antennas are measured under the industrial standard test

measurements. The project was constructed within the timeframe given. This project

will definitely prove to be a great milestone in the learning curve for the antenna manu-

facturing.

Approximately 80 hours of working hours was required for the completion of the entire

project.

It would be better if the focus would be also made on the antenna beamwidth and affect

of polarization. There are few other parameters such as directivity; Axial Ratio etc could

be included further on the study.

14

REFERENCES

[1] Balanis, C.A.: Antenna Theory - Analysis and Design (3rd Edition), available at the

TUT digital library (dLib) => ebooks => Knovel => search for “Balanis”

[2] Johnson Components SMA flange mount jack - datasheet (MOODLE)

[3] http://www.vias.org/wirelessnetw/wndw_06_05_02.html

[4] https://moodle.tut.fi/course/view.php?id=2608

[5] https://moodle.tut.fi/file.php/2608/Guided_work/ELE-

6216_Guided_Work_Instructions_v0.pdf

[6] https://moodle.tut.fi/file.php/2608/Data/rogers_4000_data.pdf

[7] https://moodle.tut.fi/file.php/2608/Data/Johnson_SMA_flange_mount_jack.pdf

[8] https://moodle.tut.fi/file.php/2608/Data/ELE-6216_Measurement_sites.pdf [9] http://dl.dropbox.com/u/15048456/load_script.m

15

APPENDIX A: SIMULATED RESULTS

Figure A.1. Smith Chart

Figure A.2. Radiation Pattern

16

APPENDIX B: MEASURED RESULTS OF AUT

Figure B.1. Efficiency of AUT

Figure B.2. E and H-Plane Gain of AUT

17

Figure B.3. Gain of AUT with frequency