Chapter 5 DESIGN AND IMPLEMENTATION OF SWASTIKA-SHAPED FREQUENCY RECONFIGURABLE ANTENNA ON FR4 SUBSTRATE

Transcription

1 Chapter 5 DESIGN AND IMPLEMENTATION OF SWASTIKA-SHAPED FREQUENCY RECONFIGURABLE ANTENNA ON FR4 SUBSTRATE The same geometrical shape of the Swastika as developed in previous chapter has been implemented on FR4 substrate in this chapter. The substrate is having thickness of 1.6mm which is higher than the Rogers RO435 used earlier in the thesis. A thicker substrate, besides being mechanically strong, will increase the radiated power, reduce conductor loss and improve impedance bandwidth. The FR4 epoxy substrate used for the antenna design is having relative permittivity as 4.4 and dielectric loss tangent of Parametric Study of Full Swastika-Shaped Microstrip Patch Antenna The effect of variations in parameters of the frequency reconfigurable Swastika-shaped patch antenna has been studied in this section. This study is different than the study conducted in the last chapter. Here, only one parameter of the full Swastika with a circle at the center (no gap in the design) has been varied and its effect has been observed. To start with, the optimized parameters for a single layer compact size Swastika-shaped frequency reconfigurable antenna for WLAN indoor and outdoor applications are presented in Table 5.1. The study is based on simulations only Study of Variation in the Radius of Center Circle (R) In this section, the radius of center circle (R) has been varied around its optimized value in the Swastika-shaped patch resonating at 5.8GHz without gaps in the design. Three cases of R equals to 7.3mm, 7.4mm and 7.5mm are studied. The return loss plots for this study are shown in Figure 5.1. It has been observed that the increase in size of the center circle decreases the resonant frequency. For the observation values, the change is not very significant. The reason behind this may be the fact that the overall size of the patch is not changed. Hence, this parameter can be used for fine tuning of the resonant frequency and impedance matching. Except for 73

2 Return Loss (db) the frequencies greater than 9GHz, all the three curves are almost same. The -1dB bandwidths obtained in these cases are 125MHz. Table 5.1 Parameters of the frequency reconfigurable Swastika-shaped antenna design Parameter Radius of center circle (R) Length of Horizontal & Vertical Stub (L p and W p ) Length of Side Stub at 45 degree (L s ) Width of the Rectangular Sections (W t ) Diameter of four Dots (D) Co-ordinates of four Dots (x d, y d ) Feed Point radial distance at 45 degree from center Feed Point co-ordinates (x f, y f ) Ground size (L g x W g ) Gap near circle for frequency reconfigurability Optimized Value 7.4 mm 32.4 mm 5 mm 2 mm 4 mm (9,9),(9,-9),(-9,-9),(-9,9) mm 2.2 mm ( , ) mm 5 mm x 5 mm 1 mm. Varying Radius 7.3(R),7.4(G),7.5(B)) Name X Y m m m m m2 m Freq [GHz] Fig. 5.1 Return loss plots for study of radius of circle in full Swastika shape 74

3 Return Loss (db) Study of Variation in the Length of the Horizontal and Vertical Strips (L p and W p ) In this section, all the parameters of full Swastika are same as shown in Table 5.1 except the Length of the horizontal and vertical Stubs (L p and W p ) which has been varied to observe the effect of overall dimension of the geometry. Three values 3.4mm, 32.4mm and 34.4mm are considered. A comparative return loss plots for these values are presented in Figure 5.2. The results are again consistent with the theory that increasing the overall size of the patch will decrease the resonant frequency. The shapes of the curves are almost same except the left side shift with increase in the size. The multiband response can be seen in the plots.. Varying stub length 15.2(R),16.2(G),17.2(B) Name X Y m m m m3 m2 m1 3.4 mm 32.4 mm 34.4 mm Freq [GHz] Fig. 5.2 Return loss plots for study of variation in the length of the horizontal and vertical strips 75

4 Return Loss (db) Study of Variation in Dot Location (±x d, ±y d ) In this section, the location of the dots, one dot in each quadrant in full Swastika design is varied and the effect has been observed. All other parameters are same as shown in Table 5.1. Four different values are considered as (±8mm, ±8mm), (±8.5mm, ±8.5mm), (±9mm, ±9mm) and (±9.5mm, ±9.5mm). The return loss plots for these values are shown in Figure 5.3. It has been noted that for first three values, the curves are overlapping except very small change in the value of return loss at some dips but for (±9.5mm, ±9.5mm), the resonant frequency has been changed considerably. To further study the effect of dots in the geometry, a comparative graph of return loss for presence and absence of dots has been presented in Figure 5.4. By addition of dots in the design, the resonant frequency has been changed from 5.5GHz to 5.8GHz. The effect is same as if the size of the patch is decreased to increase the resonant frequency. In all the cases, the return loss is always better than -22dB. The maximum return loss of -25.7dB has been obtained for (±9mm, ±9mm) case at 5.8GHz frequency. All the curves are showing multiband response. The first three cases of the study are having other resonating frequencies higher than 5.8GHz frequency and these are in the vicinity of 7.25GHz and 9.6GHz. For (±9.5mm, ±9.5mm) case, the other resonating frequencies are at 6.8GHz and near 1GHz. If not required, these bands can be easily filtered out. Ansoft Corporation XY Plot Name X Y m mm, 8mm -2. m m m m4 m1 m2 8.5mm,8.5mm 9.mm,9.mm 9.5mm,9.5mm -25. m Freq [GHz] Fig. 5.3 Return loss plots for study of dot location 76

5 Return Loss (db) Ansoft Corporation. XY Plot 3 HFSSDesign Name X Y m m Without dots With dots at 9mm, 9mm -25. m1 m Freq [GHz] Fig. 5.4 A comparative return loss plots for without dot and with dot condition at ±9mm, ±9mm locations To further investigate the effect of addition and location of the dots, the E and H field distribution of all the above mentioned cases are examined and shown in Figure 5.5 and Figure 5.6 respectively. It has been clearly visible that the location of dots affects the field distributions for both E and H. If Figure 5.5 (d) and (e) are compared, it has been noticed that in without dot case, two stubs opposite to each other are not having good field strength and if the dots are added, the electric field is almost symmetrically distributed to all four stubs. Further, it has been seen that as the dots move out, the field strength on the remaining patch is improved. It is happened because when the dots are in proximity to the center circle, they are mutually coupled to it and hence drawing the energy from the patch. When they are far from the circle, this mutual coupling is weakened and now more energy is flown into the patch. Similar is the case for H field distribution. 77

6 (a) (b) (c) (d) (e) Fig. 5.5 E-Field distribution on the Swastika geometries of different locations of dots (a) 8mm, 8mm (b) 8.5mm, 8.5mm (c) 9mm, 9mm (d) 9.5mm, 9.5mm (e) No dot 78

7 (a) (b) (c) (d) Fig. 5.6 H-Field distribution on the Swastika geometries of different locations of dots (a) 8mm, 8mm (b) 8.5mm, 8.5mm (c) 9mm, 9mm (d) 9.5mm, 9.5mm (e) No dot (e) 79

8 The far field radiation patterns in E and H plane for without dot and with dots at (±9mm, ±9mm) positions at their respective resonant frequency are shown in Figure 5.7 (a) and (b) respectively. It has further been noted that apart from symmetrical distribution of E and H fields due to the presence of dots in the Swastika design, the maximum value of gain for Φ = and 9 has been increased from.38db to.95db and from.98db to 2.dB respectively due to the presence of dots. Ansoft Corporation Radiation Pattern 3 HFSSDesign1-3 m Name Theta Ang Mag m m db(gaintotal) Freq='5.545GHz' Phi='deg' -9 m2 9 db(gaintotal) Freq='5.545GHz' Phi='9deg' Ansoft Corporation (a) Radiation Pattern 2 HFSSDesign m1 3 Name Theta Ang Mag m m db(gaintotal) Freq='5.815GHz' Phi='deg' -9 m2 9 db(gaintotal) Freq='5.815GHz' Phi='9deg' (b) Figure 5.7 E and H plane radiation patterns (a) without dots at 5.5GHz (b) with dots at ±9mm, ±9mm position at 5.8GHz 8

9 Return Loss (db) Study of Variation in Dot Diameter (D) The size of the dot has been varied in this section while keeping their position same. All the other parameters are same as given in Table 5.1 for full Swastika design. Three values of diameter of dots as 3mm, 4mm and 5mm at (±9mm, ±9mm) position in each quadrant are considered for the study. The return loss plots have been shown in Figure 5.8. It has been observed that the variation in dot size doesn t affect the resonant frequency. The shape of the return loss curves for all the three cases are exactly same up to 9GHz. Only small variation in the value of return loss at 5.8GHz is seen in the plot.. Varying dot diameter 3(R),4(G),5(B) Name X Y m m m m3 m1 m2 3mm 4mm 5mm Freq [GHz] Fig. 5.8 Return loss plots for variation in diameter of dots Study of Variation in Length of the Side Stub (L s ) The length of the side or end strip connected to each L-shaped section is varied and the effect has been observed. Three values are considered as 4.5mm, 5mm and 5.5mm. The angle of the side stub is 45 from the Y-axis. The return loss plots are shown in Figure 5.9. The increase in length of the end strip increases the overall size of the patch hence resonant frequency decreases. The resonant frequency corresponding to 4.5mm, 5mm and 5.5mm are 5.86GHz, 5.81GHz and 5.72GHz respectively. The shape of the curves is almost same for all the cases. 81

10 Return Loss (db). Varying side stub length 4.5(R),5,5(G).5(B) Name X Y m m m m3 4.5mm 5.mm -25. m2 5.5mm Freq [GHz] Fig. 5.9 Return loss plots for variation in length of side stub m Study of Angle of the Side Stub (θ S ) The angle of the side or end stub connected to each L-shaped section has been varied in this section of study. All the other parameters are same as mentioned in Table 5.1. The angle is measured from Y-axis. Three values of the angle as 3 degree, 45 degree and 6 degree are considered. The return loss plots of these three values are shown in Figure 5.1. As the angle increases, the resonant frequency also increases. Hence, it can be concluded that the effect of increasing the angle of the side stub is similar to the decrease in patch size so that the resonant frequency increases. 82

11 Return Loss (db). Varying stub angles 3(R),4(G),6(B) Name X Y m m m m1 3 degree 45 degree -25. m2 6 degree -3. m Freq [GHz] Fig. 5.1 Return loss plots for variation in angle of side stub Study of change in Feed Point in a Full Swastika Design All the study of previous cases was carried out on a feed point that is located at a distance of 2.2mm on a radial line which is at 45 degree in a quadrant. Due to symmetry of the design, the feed point may be located in any of the quadrant at the above mentioned point. The return loss plots for a feed point at the same line and a distance of 1.8mm, 1.9mm, 2.mm, 2.1mm, 2.2mm, 2.3mm and 2.4mm are shown in Figure It has been observed that the change in feed mainly affects the matching of the antenna impedance and the feed. Better the matching, higher the return loss has been obtained. For the first four cases, the resonant frequency is 5.77GHz and for remaining three, it is 5.8GHz. For 5.77GHz resonant frequency, the return loss improves continuously as the feed has been moved far from the center of the circle. As soon as the feed is moved to 2.2mm, the resonant frequency has been changed to 5.8GHz with a return loss of -25.7dB. The return loss has been decreased at the same resonant frequency if the feed point is moved farther. The feed point to be considered finally for the antenna has been decided at 2.2mm on the same line. 83

12 Return Loss (db) Return Loss (db) Ansoft Corporation. XY Plot 4 HFSSDesign Name X Y m m m m m m m m1 m2 m3 m7 m6 m5 m4 1.8mm 1.9mm 2.mm 2.1mm 2.2mm 2.3mm 2.4mm Freq [GHz] Fig Return loss plots for variation in probe feed point Study of Size of Ground in a Full Swastika Design In this section of study, the ground size has been varied and its effect has been observed. Five square sizes of side length 45mm, 5mm, 55mm, 6mm and 65mm are studied. The simulated return loss curves are shown in Figure It has been noticed that the size of ground doesn t have major effect on resonant frequency but impedance matching has been changed with the change in ground size. No effect on shape of the radiation pattern has also been observed. Ansoft Corporation. XY Plot 2 HFSSDesign Name X Y -2. m m m m1 m3 m5 45mm by 45mm 5mm by 5mm 55mm by 55mm m m m4 6mm by 6mm 65mm by 65mm -25. m Freq [GHz] Fig Return loss plots for study of size of ground 84

13 Return Loss (db) Study of Length of Gap near Center Circle for Frequency Reconfigurability The gap between center circle and the four rectangular strips to create frequency reconfigurability is primarily concerned with the size of the practical switches to be placed in the gaps. The return loss plots for three gap sizes of 1mm, 1.5mm and 2mm are shown in Figure It has been observed that the variation in gap length doesn t affect the resonant frequency but slightly changes the return loss.. XY Plot Name X Y m m m mm 1.5mm m3 m2 m1 2mm Freq [GHz] Fig Return loss plots for study of variation in length of gap near circle 85

14 5.2 Frequency Reconfigurable Antenna Design The parametric study of a single layer compact size frequency reconfigurable antenna designed for WLAN indoor and outdoor services on Fr4 substrate of thickness 1.6mm was presented in previous section. The study was conducted around the optimized parameters presented in Table 5.1 for the required reconfigurability between bands at 5.2GHz and 5.8GHz. The same parameters are used for simulations of the antenna designs with all switches ON [96] and introducing frequency reconfigurability by making all switches ON and OFF [97]. The two geometries are shown in Figure A small dot in the center circle is showing the feed point. (a) (b) Fig Antenna geometries representing (a) all switches ON (b) all switches OFF The simulated return loss plots for all switches ON and all switches OFF conditions are shown in Figure For all switches ON condition, the response is multiband having resonant frequencies at 7.25GHz and 9.7GHz in addition to the desired resonant frequency at 5.8GHz. If the higher frequency bands are not required, they can be easily filtered out. If all the switches are OFF, the center circle is detached from the four sections and the resonant frequency is moved to 5.2GHz. The effect of creating gaps in the design is somewhat similar to the effect of increasing the patch dimensions so that the resonant frequency decreases. The value of return loss obtained at 5.2GHz is -3.7dB having lower -1dB frequency at 5.128GHz and higher -1dB frequency at 5.323GHz resulting in 195MHz bandwidth. The simulated bandwidth is marginally short from the bandwidth required for IEEE82.11a indoor services. A 86

15 return loss of -25.7dB has been obtained at 5.8GHz frequency with lower -1dB frequency at 5.742GHz and upper -1dB frequency at 5.868GHz resulting in 126MHz bandwidth sufficient for IEEE82.11a indoor services. The reason of selecting full patch to resonate at 5.8GHz and the patch with gaps to resonate at 5.2GHz is obvious since the band at 5.8GHz has 125MHz bandwidth (5.725GHz-5.85GHz) and the band at 5.2GHz has 2MHz bandwidth (5.15GHz-5.25GHz). The VSWR curves for both the switching conditions are shown in Figure A VSWR value of 1.8 is obtained at 5.2GHz and at 5.8GHz, it is 1.5 showing good matching of the antennas. The E and H plane far field radiation patterns at 5.8GHz frequency for all switches ON and at 5.23GHz frequency for all switches OFF condition are shown in Figure The E plane patterns for both the switching conditions are almost same in shape and the direction of maximum radiation is also same except its magnitude. The direction of maximum radiation for H plane has been changed but it has been observed that the shapes of the patterns are not changed much at two different frequencies. The E and H field distributions along the patch surface in both the switching conditions are presented in Figure 5.18 and Figure It has been again observed that not much energy is radiating through four dots for the similar reason discussed in previous chapter. In Figure 5.18, it has been observed that the strength of electric field at the sharp corners and bands is less in comparison to other areas whereas the strength of magnetic field is lesser in the areas near to the dots. The effect of small gaps of 1mm can be seen in Figure 5.19 as some energy has been observed in the areas nearer to the circle due to mutual coupling. Figure 5.2 shows the 3D polar plots for radiation patterns for without and with gaps in the design. These are almost same in the shape. The Smith chart for impedance variations of both the designs from 4GHz to 7GHz are shown in Figure For full swastika without gap design, the antenna reactance is inductive at most 87

16 VSWR RETURN LOSS (db) of the frequency points but at few points, it is capacitive also. In with gap design, the antenna impedance is inductive throughout the observation frequency range. The Gain v/s Frequency curves for both switching conditions for Φ = and 9 values and at θ = are shown in Figure The curves for Φ = and 9 values are exactly same for both the switching conditions. It has been observed that gain has been obtained from nearly 4GHz to 9.6GHz frequency range in all switches OFF condition whereas with all switches ON, the Gain has been observed in the vicinity of 6GHz only. The other parameters such as peak directivity, peak gain, and radiation efficiency are shown in Table 5.2. The value of Axial Ratio is almost zero indicating that it is a linearly polarized antenna. Ansoft Corporation. XY Plot 1 HFSSDesign Name X Y m m All Sw itches OFF All Sw itches ON -25. m2-3. m Freq [GHz] Ansoft Corporation 2. Fig Simulated return loss plots for all switches OFF and all switches ON XY Plot 3 HFSSDesign1 15. All Switches OFF All Switches ON Freq [GHz] MX1: MX2: Fig Simulated VSWR plots for all switches OFF and all switches 88

17 Ansoft Corporation Radiation Pattern 5 Ansoft Corporation HFSSDesign1 Radiation Pattern1 m1m m Name Theta Ang Mag -3.2 Name Theta Ang Mag -.2 m m m m m db(gaintotal) Freq='5.815GHz' Phi='deg' db(gaintotal) Freq='5.815GHz' Phi='9deg' db(gaintotal) Freq='5.23GHz' Phi='deg' db(gaintotal) Freq='5.23GHz' Phi='9deg' (a) (b) Fig E and H plane radiation patterns for (a) all switches ON (b) all switches OFF (a) (b) Fig E-Field distribution on patch surface for (a) all switches ON (b) all switches OFF (a) Fig H-Field distribution on patch surface for (a) all switches ON (b) all switches OFF (b) 89

18 GainTotal (db) (a) (b) Fig D polar plots for (a) all switches ON (b) all switches OFF Ansoft Corporation Smith Plot 1 Smith Plot 1 HFSSDesign St(Cylinder2_1_1_1_T1,Cylinder2_1_1_1_T1) Setup1 : Sw 12 eep St(Cylinder2 Setup1 : Sw eep (a) (b) Fig Smith charts for 4-7 GHz frequency (a) all switches ON (b) all switches OFF Ansoft Corporation 5. XY Plot 7 HFSSDesign db(gaintotal) All Sw itches OFF Phi='deg' Theta='deg' db(gaintotal) All Sw itches OFF Phi='9deg' Theta='deg' db(gaintotal) for All Sw ithes ON Phi='deg' Theta='deg' db(gaintotal) for All Sw ithes ON Phi='9deg' Theta='deg' Freq [GHz] Fig Simulated Gain v/s Frequency curves for both conditions of switching 9

19 Table 5.2 Simulated results for Swastika-shaped frequency reconfigurable antenna Parameter All Switches ON All Switches OFF f r = 5.81GHz f r = 5.23GHz Peak Directivity Peak Gain Radiation Efficiency Fabrication of Antennas FR4 epoxy of 1.6mm thickness, relative permittivity (ε r ) of 4.4 and dielectric loss tangent (tanδ) of.2 has been used as substrate material for the fabrication of the antennas. Though the loss tangent of FR4 is high resulting in low gain but easy availability and cost are also considerable factors. The patch dimensions are shown in Table 5.1. Two prototypes (with and without conductors representing the ideal ON and OFF condition of switches respectively) are fabricated using standard silk screen printing method of PCB fabrication. Double sided copper layer substrate was used. On one side, the patch was printed and the other side has been used as ground plane. After preparation of the prototype, SMA connector was soldered to the bottom side of the substrate for probe feeding. These fabricated antennas are shown in Figure 5.23 and Figure Figure 5.25 shows the fabricated antenna with four PIN diodes BAR63-3W as switches. The data sheet of this diode may be obtained from [98]. When the diodes are forward biased, they act as short circuit and when the diodes are reverse biased or unbiased, they act as open circuit. A simple biasing arrangement of the PIN diodes is shown in Figure The cathode of each diode has been connected to the ground through 1KΩ resistor. The anode of each diode is connected to the center circle which is further connected to one resistor of 1KΩ. To make diodes ON, a +5V supply has been given to anodes through 1KΩ resistor and circular patch while ground has been connected to all four cathodes through 1KΩ resistors. The diodes are OFF as soon as the +5V supply is removed. 91

20 (a) (b) Fig Photograph of fabricated antenna showing conductor for all switches ON condition (a) Top side showing patch (b) Bottom side showing ground plane and SMA connector (a) Fig Photograph of fabricated antenna showing gaps for all switches OFF condition (a) Top side showing patch (b) Bottom side showing ground plane and SMA connector (b) 92

21 Fig Photograph of fabricated antenna with PIN diode BAR63. The top side is showing patch, four diodes and bias circuit Fig A simple biasing arrangement for BAR63 PIN diodes on the patch. The value of resistors is 1KΩ. 5.4 Comparison of Simulated and Measured Results The measurements for return loss and E and H plane radiation patterns for Gain Total of the fabricated antennas were carried out with the same measurement setup described in chapter 4. The measured and simulated return loss plots (S11 93

22 Return Loss (db) parameter) for full Swastika patch (all switches ON condition represented by the presence of conductor) are shown in Figure The shapes of the curves are similar but shifting of the curve at some dips are seen. The comparison of simulated and measured results of return loss has been presented in Table 5.3. The measured and simulated return loss plots (S11 parameter) for full Swastika patch with gaps (all switches OFF condition represented by the absence of conductor) are shown in Figure The measured curve is just following the simulated curve with very small shift. The comparison of simulated and measured results of return loss in all switches OFF condition has been presented in Table 5.4. Similar to the results obtained for the prototype representing all switches ON condition, here also, the measured bandwidth is more than the simulated bandwidth Simulated Measured Frequency (GHz) Fig Return loss plots for all switches ON condition Table 5.3 Simulated and measured results of return loss plots for all switches ON Parameter All Switches ON Simulated Measured Resonant Frequency (f r ) 5.81GHz 5.81GHz Return Loss at f r -25.7dB dB Lower -1dB Frequency 5.742GHz 5.725GHz Upper -1dB Frequency 5.868GHz 5.875GHz Bandwidth 126MHz 15MHz 94

23 Return Loss (db) Simulated Measured Frequency (GHz) Fig Return loss plots for all switches OFF condition Table 5.4 Simulated and measured results of return loss plots for all switches OFF Parameter All Switches OFF Simulated Measured Resonant Frequency (f r ) 5.23GHz 5.275GHz Return Loss at f r -3.7dB -22.9dB Lower -1dB Frequency 5.128GHz 5.14GHz Upper -1dB Frequency 5.323GHz 5.38GHz Bandwidth 195MHz 24MHz The simulated and measured E and H plane radiation patterns for both the prototypes are compared in Figures 5.29 to It may be noted that all the measured patterns are in good agreement with simulated patterns except some negligible errors at few points. The measured magnitude and direction of maximum gain is almost equal to the simulated results. For all switches ON condition, the measured maximum Gain Total at Theta equals to -5 degree for E plane is 1.dB in comparison to simulated gain of.96db and for H plane, the measured gain is 1.7dB in comparison to the simulated value of 2.dB at Theta equals to 9 degree. For all switches OFF condition, the measured maximum Gain Total for E plane at Theta equals to -5 degree is 3.8dB in comparison to 3.8dB gain at Theta equals to -4 degree and for H plane, the 95

24 measured gain is 3.8dB at Theta equals to degree in comparison to 3.8dB simulated value at Theta equals to -2 degree. The results of gain in H plane for all switches ON condition are having more deviation from simulated results than the other results Measured Simulated Fig Simulated and measured E plane radiation patterns at 5.8 GHz for all switches ON condition Measured Simulated Fig. 5.3 Simulated and measured H plane radiation patterns at 5.8 GHz for all switches ON condition 96

25 Measured Simulated Fig Simulated and measured E plane radiation patterns at 5.25 GHz for all switches OFF condition Measured Simulated Fig Simulated and measured H plane radiation patterns at 5.25 GHz for all switches OFF condition 97

26 The antenna fabricated using PIN diodes BAR63 shown in Figure 5.25 was tested for return loss parameter using HP s spectrum analyzer model number 8559A having operating frequency from.1 to 21GHz [99-1]. The measurement set up and results can be seen in Figure 5.33 and When the diodes are not biased, a return loss of about -13dB has been obtained at 5.37GHz which is very less in comparison to the simulated value. When the diodes are forward biased by giving 5V DC supply, the diodes are ON and a return loss of about -22dB has been observed at 6.GHz resonant frequency. It has been observed that the resonant frequency has been shifted in both the cases when the diodes are connected. Fig Return loss measurement when diodes are OFF Fig Return loss measurement when diodes are ON 98

27 5.5 Summary A single layer, compact size Swastika-shaped frequency reconfigurable microstrip patch antenna for WLAN indoor and outdoor applications has been studied, designed and fabricated in this chapter. In the study part, all the parameters of the design have been varied one by one around the optimized value while keeping all the other parameters at their optimized value. The normal form of Swastika without four dots was simulated and all its relevant parameters were detected which revealed that the electric field distribution was asymmetric in nature. After adding four dots, one in each quarter, the distribution of electric field became symmetric. The addition of four dots and inclined sections at the end of L-shaped section represents the most standard form of Swastika being used at different occasions. The antenna has been designed to operate at 5.2GHz when all the switches are OFF and at 5.8GHz when all the switches are ON. The switches are assumed to be ideal and two prototypes are developed. The dimension of the antenna is 5 5mm 2. The measurements of return loss (S11 parameter) and the E and H plane radiation patterns are carried out using industry standard measurement set up. The proposed antennas have simulated return loss of -3.7dB at 5.23GHz and -25.7dB at 5.81GHz. The simulated bandwidth in all switches ON and all switches OFF conditions are 126MHz and 195MHz respectively. The fabricated prototype antennas have measured return loss of -22.9dB at 5.275GHz and dB at 5.81GHz. The measured bandwidth in all switches ON and all switches OFF conditions are 15MHz and 24MHz respectively. A good agreement between simulated and measured return loss, E plane and H plane radiation patterns have been observed for the two prototypes. The simulated peak gains of 2.43dB and 1.6dB are observed at 5.2GHz and 5.8GHz resonant frequency respectively. The low gain is the result of substrate properties. As an experiment, the BAR63-3W PIN diodes are placed on the antenna for the switching purposes. When the diodes are ON, a return loss of -22dB has been observed at 6.GHz resonant frequency and when the diodes are unbiased, the resonant frequency is 5.37GHz with return loss of -13dB. Though the data sheet recommends the diodes to be used up to 3GHz only because of increased insertion 99

28 loss but it was an effort to understand the effect of using practical switches in the microstrip patch antennas. The variations in the results may be due to the characteristics of diodes and the biasing wires present on the antenna structure. The use of switches which do not need bias lines and circuit is very important for reconfigurable antennas because the presence of additional lines or wires on the radiator degrades the radiation performance of any antenna since the RF leakage through the bias lines cannot be avoided. The measured results verify the good performance of proposed antenna but due to the substrate properties, low gain performance of the proposed antenna has been observed. 1