You are on page 1of 19


P University of Engineering and Technology Mardan Campus Department of Telecommunication Engineering Term Project Report

Project Title Design of dual band patch antenna




com Project coordinator Dr. Professor Telecommunication Engineering             . Sadiqullah Advisor.haseebak@gmail.           Project By: Haseeb Ahmed Khan Email id: engr.

It has a ground plane with the dimension of w*L*H= 255*255*2.45 GHz frequency band the antenna is detuned and not matched. The patch has also truncated corner. The simulated antenna was detuned in frequency so it operates best for 828 MHz with circular polarization straight out from the patch and linear polarization at the short sides (circular close to antenna). The disadvantage with this feeding technique is the manufacturing process.5 mm. To simplify the manufacturing process of the antenna truncated corner are used instead of the dual feed patch. Therefore a feeding technique that feed the antenna below the patch will be the best alternative since it decrease the size (width or length) of the antenna. Since the reader antenna that is used in tests is right handed circular polarized the designed antenna also need to be right handed circular polarized for low polarization loss factor as possible.287 GHz the RL< -6 db. which means they can not communicate with each other.Starting and simulation stage: First it needs to choose which feeding technique that will be used. . Between 1. Example of these feeding techniques is coaxial probe or coaxial probe with capacitive feed. It can be difficult to match the antenna. For the 2.45 GHz the antenna should radiate from the short sides of the antenna.5 mm and with patch dimension W*L*H= 153*153*2. When feed and polarization technique have been chosen the design continuous with a already existing large patch antenna for 868 MHz. For 2. A microstrip line is not the best choice since the microstrip line increase the size of the antenna and the bandwidth decrease if a thick substrate is used. The used feed technique is coaxial probe with capacitive feed with air substrate of height 13 mm.9 GHz and 2. If the designed antenna is LHCP and reader antenna is RHCP the polarization loss factor will be maximal. A small size and wide bandwidth are two properties with high priority.15 GHz at the sides of the antenna. Simulations result for the antenna shows that it has a bandwidth of 90 MHz (810-890 MHz) with RL < -10 db. An aperture-coupled patch would not either be the best solution since it gives narrow bandwidth and difficult to fabricate. Than the choice will be coaxial probe with capacitive feed. The coaxial probe gives narrow bandwidth and coaxial probe with capacitive feed gives wider bandwidth. The simulated antenna radiates with around 5 dBi at 2.

FR-4 is a very common material that is cheaper2 than TMM 10i and with Er= 4.45 GHz frequency band the antenna is detuned and not matched. To get a height of 12-13 mm several layers need to be stacked.Minimizing stage: To reduce the size of the antenna three techniques been used: shorting wall. The substrate (TMM 10i) has a delivery time of 7-8 weeks which is not acceptable for the current project. 2.5mm). The minimum bandwidth that is needed is at least 2 MHz bandwidth for the antenna. L > W.7 db and for 2.21 mm) for the antenna. There occurs one big problem with the minimized antenna. The antenna will also get higher bandwidth with FR4 instead of TMM 10i since it has lower dielectric constant gives also a recommendation for the width of the patch.02.8-1. For the 2.75 GHz the RL< -4. To minimize the total physical size of the antenna the ground plane is kept as small as possible.1.3 and tan theta = 0. FR-4 is only available in few thicknesses (0. Therefore the substrate is changed to FR-4.8 and tan theta 0.002 . The advantage with this substrate is that Rogers Corporation has many different heights of the substrate available. That is an effective method and reduced the size of the patch with 50%. The reason why the extension part (6h antenna) is so long depends on the large thickness of the antenna. By using this method to reduce the size of the antenna the bandwidth will decrease since the bandwidth is dependent of the ratio between the width and length of the patch.5 mm. To increase the frequency the antennas size can then be minimized. The consequence with FR-4 instead of TM10i is that FR-4 has lower dielectric constant which increases the size of the patch antenna (formulas 2.6). This will increase the total physical size of the antenna significantly. The bandwidth for 868 MHz is 24 MHz (856-880 MHz) for RL< -10 db. RL868 MHz= -42 db. In this design the ratio between the length and width of the patch is exactly two. The minimized antenna is matched for 868 MHz. The goal was to get higher than 2 MHz bandwidth to compensate detuning changes of frequency. First a shorting wall is used to minimize the antenna. A substrate with higher dielectric constant lowers the resonance frequency but it will also decrease the bandwidth. A substrate with high dielectric constant is Rogers Corporation TMM 10i. change of substrate and use of available antenna parameters. The antenna was simulated with TMM 10i which gave a resonance frequency lower than 868 MHz. The ground plane affects the size of the antenna. Er = 9. A 12. Therefore it was necessary to change substrate or design. To decide the size of the .138 GHz the RL< -10.2. The recommended size of the ground plane will not be used. The antenna dimension after minimizing it with TMM 10i is W*L*H= 108*52*3.8 db. 2. Air is an excellent choice as substrate if high bandwidth is a priority. By reducing W the bandwidth decreases. The whole project time is 20 weeks and when the materials are delivered the project would nearly be ended. At 2.7 mm thick substrate is needed. The ground plane should be extended with the factor 6h antenna (81.

red curve in figure 4.ground plane. In CST the patch circle was not centered over the feed circle which it were drilled and for CST the radius was 3 mm. that is 42 MHz difference. Simulation shows that the built antenna is better match then the simulation for the design with errors but there are still 44 MHz difference between the built antennas S11 parameter. 16 MHz bandwidth for RL < -10dB but the network analyzer showed that the antenna is matched for the resonance frequency 880 MHz (around 10 mm from metal). not diameter of 3 mm. When the antenna was built the network analyzer3 showed that the antenna has a resonance frequency around 650 MHz and 1.5 mm) the design get more stable and easier to build. To simplify the manufacturing process. simulation result and the manufacturing process has been taking into account. Design errors: From the beginning the simulated model of the antenna for 868 MHz has a bandwidth about 25 MHz for RL < -10 dB. The final test bed needed two RFID antennas. the choosing length of the antenna is 160 mm. Then the ground plane is larger than the patch and will not increase the resonate frequency as much as possible. The simulation software (CST) showed that the antenna is matched for the resonant frequency at 838 MHz (about 10 mm from metal).2. After these changes the antenna works much better and increase the frequency closer to 868 MHz but the antenna did not work as the simulations.5 mm to 1. By change the dimensions of the patch. see figure 5.12a and the simulated version. Each FR-4 plate that is bought is 160 mm long. . The width is determined by the simulation results so the antenna is match for the right frequency. When the second antenna was built few errors was found. radius of feed circle (from 3 mm to 10 mm) and radius of the hole in the patch (from 0.3 GHz.

The antenna is also built with the given parameters in figure 4.3-4.7 shows the structure from five different views of the antenna.3-4-7. .Layout: Figure 4.


Figure 4.7 shows the antenna from the short side.5 shows the patch layers form with the substrate behind.4 shows the top view from ground.5 mm height each. 1. Figure 4. . The substrate is divided into nine sub layers (SL) as figure 4. From this view only the ground is shown and the coaxial feed in the middle of the ground.Figure 4.6 shows the antenna form the long side. Figure 4.6. The antenna is built with nine layers FR-4.3 shows the top view on feed with substrate behind. Figure 4.


7-2. The simulation settings in CST (mesh shells properties.22-2.) affect the simulations result a little.45 GHz. Following simulation settings have been used: Frequency range: 800-870 MHz. There is a different between the simulated result and the measured result from the network analyzer. The simulation results are for the built antenna dimensions. Therefore the S11 parameter looks a bit different depending on which settings are used. far field pattern and axial ratio for both frequency band.General simulations settings in CST: CST microwave studio is used as simulation software. smith chart. . 868 MHz and 2. accuracy etc. 0. frequency range.5 GHz. 2.5 GHz Solver settings (Accuracy): -30dB Global Mesh Properties • Lines per wavelength: 15 • Lower mesh limit: 5 • Mesh line ratio limit: 20 Simulations result: Following section presents simulation results from CST for the S11 parameter. special the S11 parameter.

1 shows that the antenna is not matched for 868 MHz.2 % of the incident power is reflected. . RL868 MHz = −2.S11 parameter Figure 5.06 dB gives that 62.6 % of the incident power is reflected. RL836 MHz = −22 dB gives that 0.

.2 shows that the antenna is not match for 2.5 dB gives that 0.3 % of the incident power is reflected.402 GHz = −2. RL2.284 GHz = −22.Figure 5. RL2.06 dB gives that 62.2 % of the incident power is reflected.45 GHz.

3 dB gives that 0.3 shows the S11 parameter with both frequency bands (868 MHz and 2. RL868 MHz = −3.3 % of the incident power is reflected. .4 % of the incident power is reflected.5 dB gives that 44.284 GHz = −25..Figure 5.7 % of the incident power is reflected RL2.402 GHz = −1.96 dB gives that 63.45 GHz) in the same figure.7 % of the incident power is reflected.3 dB gives that 7. RL836 MHz = −11. RL2.

4 shows that the antenna is good matched at 836 MHz since the blue point is near ZS =1 and worse matched at 868 MHz since the black point is far fromZS =1.284 GHz since the blue point is near Z S = 1 and worse matched at 2.4 shows even that Z A = 46 + 6.5 shows that the antenna is good matched at 2. Figure 5. Z A = 6.66i Ω Figure 5.402 GHz since the black .Smith chart: Figure 5.25i Ω at 836 MHz and for 868 MHz.74 + 18.

.4844 dB. the total efficiency is 4.708 dB and the directivity is 5.924 dBi. Z A = 165.284GHzand for 2. For 868 MHz there are much losses in the antenna since the ratio between the radiation efficiency and total efficiency is not near 1.0.1i Ω Far field pattern: Figure 5.5iΩ at2.Figure5.6 shows the far field pattern for 868 MHz.5showseventhat ZA =56−2. The radiation efficiency is .92 dBi in front of the antenna and worse at the sides (0 dbi). The antenna radiates with 5.402 GHz.9 + 203.pointisfarfrom ZS =1.

8 shows that the antenna radiates with 7.287 dB. the total .7-5.63 dBi at best from the both short sides of the antenna for 2.Figure 5. The radiation efficiency is -2.402 GHz.

efficiency is -6.605 dB and the directivity is 7.631 dBi. Figure 5.shows that the antenna has much better circular polarization for 836 MHz than 868 MHz since the axial ratio value (dB) is closer to zero and it has also lower value for axial ratio at more angels in the far field pattern.9a and 5.402 GHz there are much losses in the antenna since the ratio between the radiation efficiency and total efficiency is not near 1. A result of this is shown in figure 5. . For 2.9b.9 a-b. Axial ratio: The simulation results show that the antenna is much better matched for 836 MHz than 868 MHz.

Figure 5.11 show that the antenna has .Figure 5.10 that the antenna has circular polarization at the left side close to the antenna and linear polarization far away from the antenna.

not so good circular polarization at the right side close to the antenna and linear polarization far away from the antenna. .