Antenna Theory and Design Project

UNIVERSITY OF ARIZONA: ECE 584

Antenna Theory and Design 2013 Take Home Project Ryan Sessions 4/23/2013

“The goal of this assignment is two-fold. The first is the analysis of the performance of antenna arrays, (Steps 1-5) and the second is to design a low-profile (Planar Inverted F Antenna (PIFA)-like) antenna using Ansys’ High Frequency Structure Simulator, HFSS (Step 6).” [1]

Antenna Theory and Design: 2013 Take Home Project

Contents I. Introduction ............................................................................................................................................... 3 II. Antenna Arrays ......................................................................................................................................... 4 III. A Simulation of a Uniform Linearly Spaced Array .................................................................................... 7 IV. A Linearly Tapered Antenna Distribution .............................................................................................. 10 V. Electrically Scanned Difference Patterns ................................................................................................ 18 VI. Planar Inverted F Antennas ................................................................................................................... 25 IX. Conclusion .............................................................................................................................................. 36 X. References .............................................................................................................................................. 37 XI. Appendix A: MATLAB Simulation of A Linear Antenna Array ................................................................ 38

Figure 1: A Uniformly Spaced Linear Array ................................................................................................... 4 Figure 2: Simulation Program Flow Chart ..................................................................................................... 6 Figure 3: 6 Element Antenna Pattern Comparison- Approximate Theoretical Result vs. Simulation Result 7 Figure 4: 14 Element Antenna Pattern Comparison- Approximate Theoretical Result vs. Simulation Result ...................................................................................................................................................................... 8 Figure 5: A Comparison Between a 14 Element ULA and a 14 Element Linear Tapered Array (amin = 0.1) 11 Figure 6: Effect of Taper Parameter amin on Maximum Sidelobe Level ...................................................... 12 Figure 7: Effect of Taper Parameter amin on HPBW..................................................................................... 13 Figure 8: Effect of Taper Parameter amin on Directivity .............................................................................. 14 Figure 9: Comparison Between the Performance of a 14 Element ULA and a 14 Element Tapered Array (amin = 0.5) ................................................................................................................................................... 15 Figure 10: Grating Lobe for a 14 Element ULA ........................................................................................... 16 Figure 11: Grating Lobe for a 14 Element Tapered Array (amin = 0.5) ......................................................... 17 Figure 12: ULA Difference Array Configuration .......................................................................................... 18 Figure 13: Antenna Excitation Distribution for a 14 Element ULA.............................................................. 20 Figure 14: Difference Pattern for a 14 Element ULA (d = λ/2) .................................................................... 21 Figure 15: Difference Pattern for a 14 Element ULA (d = λ/2) With Central Null Shifted to 120° .............. 22 Figure 16: Element Distribution for the 14 Element Tapered Difference Pattern...................................... 23 Figure 17: The 14 Element Tapered Difference Pattern ............................................................................. 23 Figure 18: The 14 Element Tapered Difference Pattern (shifted to 120°) .................................................. 24 Figure 19: The Planar Inverted F Antenna (PIFA) ........................................................................................ 25 Figure 20: Multiband PIFA Configurations Studied..................................................................................... 25 Figure 21: Performance of the No Slot Case ............................................................................................... 26 Figure 22: Performance of the Control Case .............................................................................................. 26 Figure 23: Performance of the Shifted Slot Case ........................................................................................ 27 Figure 24: Performance of the Small Slot Case ........................................................................................... 27 Figure 25: Performance of the Large Slot Case........................................................................................... 28 1|Page

Antenna Theory and Design: 2013 Take Home Project Figure 26: Original PIFA from which all PIFAs in this Paper Are Scaled [3] ................................................. 29 Figure 27: Scale Factor Search Resulting in 780MHz Primary Operating Frequency ................................. 30 Figure 28: Performance of PIFA for SF = 0.8642 ......................................................................................... 30 Figure 29: First Attempt at a Dual Band PIFA Performance Results .......................................................... 31 Figure 30: Simulation Results are Used to Adjust the Frequency of Second Resonance .......................... 32 Figure 31: Performance Characteristics of First Dual Band PIFA Design .................................................... 32 Figure 32: Performance Characteristics of Second Dual Band PIFA Design................................................ 33 Figure 33: Dual Band PIFA Dimension Definition ........................................................................................ 34

Table 1: Comparison Between the Results of the Simulation and Theoretical Predictions ......................... 9 Table 2: Comparison Between the Results of the Simulation and Closed Form Numerical Predictions ...... 9 Table 3: Comparison Between the Performance of a 14 Element ULA and a 14 Element Tapered Array (amin = 0.5) ................................................................................................................................................... 15 Table 4: Parametric Study of the Effects of a Slot on the Performance of a PIFA ..................................... 29 Table 5: PIFA Dimensions, Including a PIFA Dimension Calculator ............................................................. 34

2|Page


I. Introduction The impact of the antenna on the world can hardly be overlooked. One need only examine the recent surge in wireless enabled devices to note the prevalence of antennas. Indeed, advances in antenna design have enabled vast improvements in durability and packaging of such devices, partly due to novel antenna designs such as the Planar Inverted F Antenna (PIFA). Like many recent advances, the surge in recent developments in antenna design has been facilitated by the availability of quality simulation tools to design engineers. This paper will examine several topics utilizing simulation techniques in antenna theory. The first is that of the antenna array- a configuration in which multiple antennas are combined together in order to greatly enhance the directivity with which a signal can be transmitted or received. A simulation of an antenna array is presented and the results compared to theoretical closed form predictions. The simulation is then used to synthesize an antenna array pattern from requirements. A difference antenna array pattern is synthesized, then, through the application of phase shifts, the pattern is scanned angularly. The second topic that this paper will discuss is adaptation of an existing PIFA design to enable single and dual band performance in the 4G LTE band 14 frequency range. A parametric study on dual band antennas is performed and the results are applied to a design problem along with the help of a full wave simulation tool.

3|Page


II. Antenna Arrays In order to develop a simulation of the linear array, one must first understand the theory behind an antenna array. An antenna array is composed of N individual radiators. The radiators are assumed to radiate with equal strength in all directions. A schematic depiction of an antenna array is shown in figure 1. The antenna elements are assumed to be distributed equally along a straight line with element spacing d, and relative phase β. The superposition of the waves emitted by the antennas is simply the sum of the individual phasor representation of each radiator. This superposition is known as the array factor (AF), and represents the total wave from all radiators when seen at a large distance: ∑

(

)

( ) Where the variable ψ, which represents the individual phase of a wave from each element, has been introduced. The variable k is defined to be the wavenumber of the emitted wave.

Figure 1: A Uniformly Spaced Linear Array

For the special case when the array elements all have the same output power, the an coefficients above may be assumed to be equal to 1. This special case is known as that of the Uniform Linear Array (ULA). The array factor of the ULA can be reduced through summation of geometric series to: (

)

(

⁄ )

(

⁄ )

In cases of sufficiently large arrays, (generally greater than 10), the array factor can be further simplified by the approximation: 4|Page


(

⁄ )

(

)

⁄

The resulting patterns can be shown [2] to reach half of the maximum power at the angle:

[

(

)⁄ ]

(

)⁄

One may find the absolute maximum of the array factor at the angle [2]: (

)

Therefore, one can find the angle of the beam which yields at least 50% of the maximum power within the angle known as the half power beam width, HPBW as [2]: |

|

Due to the damped oscillatory nature of the array factor, many times there will be a local maximum at some angle of power that is less than the maximum power. This is known as a side lobe. The side lobe angle may be found by [2]:

{

[

(

) ]⁄ }

(

[

) ]⁄

where s has been defined to serve as an index to distinguish individual side lobes. If one assumes a large array, then the side lobe levels (SLL) of individual sidelobes can be found by [2]: | Finally, it is often useful to have a means for quantifying how concentrated the radiation output of an array is. It is common for arrays to direct most of their energy in one general direction. In order to measure this, one need only compare the maximum radiation intensity of the array with the average radiation intensity:

∯

∯

( )

For sufficiently large ULAs, the directivity may be approximated by [2]:

5|Page

Antenna Theory and Design: 2013 Take Home Project ( ⁄ ) Using the above relations, a numerical simulation was developed in MATLAB. See Appendix A for source code. A flow chart of basic operation for the simulation is shown in figure 2 below.

Figure 2: Simulation Program Flow Chart

6|Page


III. A Simulation of a Uniform Linearly Spaced Array In order to demonstrate the accuracy of the ULA simulation program, the program results will be compared to the closed form and approximate results given in section II. The metrics that will be used are the directivity, the peak sidelobe levels, and half power beam width (HPBW). Two test cases will be used: a 14 element array and a 6 element array. The simulations were run assuming equal element excitation amplitudes and no inter-element phase differences. As can be clearly seen in figures 3 and 4, the results match for results around 90° which corresponds to ψ=0. As θ deviates from 90°, ψ deviates from 0. Recall the approximation which yielded the approximate AF formula: (

)

(

⁄ )

(

⁄ )

⁄ )

( ⁄

It is clear that this approximation is only valid for small values of ψ. As ψ increases, N ψ/2 becomes larger than sin(N ψ/2). This means that the approximate AF will get smaller than the actual AF as the difference increases, i.e. as θ deviates away from 0°, which is exactly what is shown in figures 3 and 4.

Figure 3: 6 Element Antenna Pattern Comparison- Approximate Theoretical Result vs. Simulation Result

7|Page


Figure 4: 14 Element Antenna Pattern Comparison- Approximate Theoretical Result vs. Simulation Result

A comparison between the results of the simulation program and the theoretical predictions are shown in table 1 below. When the numerical calculation that performs the summation of the phasors is replaced by the closed form solution: (

)

(

⁄ )

(

⁄ )

The results in table 2 are obtained. The reader can clearly see that the closed form solution provides excellent support for the results obtained from the simulation. Theoretical θ(°) SLL (dB)

SLL 14 Element

8|Page

1 2 3 4

77.6 69.1 60.0 50.0

-13.5 -17.9 -20.8 -23.0

Simulation Angle SLL (dB) 78.2 69.4 60.2 50.2

-13.1 -17.4 -19.9 -21.5


5 6

38.2 21.8

-24.8 -26.2

38.3 21.8

-22.4 -22.9

HPBW

7.3

7.3

Directivity

14.0

14.0

Theoretical

Simulation

SLL 6 Element

θ(°) SLL (dB) Angle SLL (dB) 1 60.0 -13.5 61.2 -12.4 2

HPBW Directivity

33.6

-17.9

34.1

-15.3

16.9

17.2

6

6.0

Table 1: Comparison Between the Results of the Simulation and Theoretical Predictions

SLL 1 2 3 4 5 6

14 Element

HPBW Directivity

6 Element

Theoretical Simulation θ(°) SLL (dB) Angle SLL (dB) 78.2 -13.1 78.2 -13.1 69.4 -17.4 69.4 -17.4 60.2 -19.9 60.2 -19.9 50.2 -21.5 50.2 -21.5 38.3 -22.4 38.3 -22.4 21.8 -22.9 21.8 -22.9 7.3 7.3 14.0

14.0

Theoretical Simulation SLL θ(°) SLL (dB) Angle SLL (dB) 1 61.2 12.4 61.2 -12.4 2 34.1 15.3 34.1 -15.3 HPBW Directivity

17.2

17.2

6

6.0

Table 2: Comparison Between the Results of the Simulation and Closed Form Numerical Predictions

9|Page


IV. A Linearly Tapered Antenna Distribution The performance of a linear array may be improved by adding a taper to the distribution. This has the effect of reducing the sidelobe levels. Unfortunately, this comes at the price of a broader main beam. Many methods exist to allow a designer to allocate the necessary excitation to each array element. Perhaps the simplest however, is that of the linear taper. The excitation is decreased linearly from the centermost elements of the array to the outermost elements of the array, where it takes the value amin. This method is straightforward enough that a designer with simulation tools (like the afore presented simulation program) may easily determine the necessary parameters of the taper by trial and error, provided of course, that the desired sidelobe levels are sufficiently above levels where only the most sophisticated methods will provide a good design. To illustrate this, an example design will be synthesized so that the 1st sidelobe level is below 18dB. A parametric study will be made of the taper parameters and the resulting sidelobe levels. The taper is assumed to have the form: (

{ (

) ⁄

⁄ )

⁄

A 14 element, d = λ/2 example of this distribution is plotted in figure 5. A variety of values between 0
10 | P a g e


Figure 5: A Comparison Between a 14 Element ULA and a 14 Element Linear Tapered Array (amin = 0.1)

11 | P a g e


Figure 6: Effect of Taper Parameter amin on Maximum Sidelobe Level

12 | P a g e


Figure 7: Effect of Taper Parameter amin on HPBW

13 | P a g e


Figure 8: Effect of Taper Parameter amin on Directivity

The effect of the taper parameter on the half power beam width and the directivity was also studied and is shown in figure 7. The HPBW was found to decrease with increasing amin. This stands to reason since the case of amin = 1 corresponds to a ULA. The directivity was found to increase as the taper parameter amin increase to 1 (see figure 8). Based on a design requirement of -20dB maximum SLL, the value amin = 0.5 was selected. The tapered distribution parameters are compared to an analogous ULA in table 3. The directivity is found to suffer some from the taper, as does the HPBW, but not severely. All sidelobe levels were found to be decreased by 3-7dB when a taper was implemented.

SLL 14 Element

14 | P a g e

1 2 3 4 5 6

θ(°) 78.2 69.4 60.2 50.2 38.3 21.8

ULA Tapered (amin = 0.5) SLL (dB) Angle SLL (dB) -13.1 77.2 -20.1 -17.4 69.1 -20.5 -19.9 59.9 -25.0 -21.5 50.0 -25.3 -22.4 38.2 -27.0 -22.9 21.8 -27.1


HPBW

7.3

8.1

Directivity

14.0

13.3

Table 3: Comparison Between the Performance of a 14 Element ULA and a 14 Element Tapered Array (a min = 0.5)

Figure 9: Comparison Between the Performance of a 14 Element ULA and a 14 Element Tapered Array (a min = 0.5)

Finally, a parametric study was made of the impact of increasing the value of d from λ/2 to slightly greater than λ. It was discovered that grating lobes, a case where sidelobes occur with the same amplitude as the main beam, were introduced as soon as d became greater than or equal to λ. This is illustrated in figures 10 and 11 for both the ULA and the tapered array.

15 | P a g e


Figure 10: Grating Lobe for a 14 Element ULA

16 | P a g e


Figure 11: Grating Lobe for a 14 Element Tapered Array (amin = 0.5)

Figure 12: Antenna Excitation Distribution for Lobe for a 14 Element Tapered Array (amin = 0.5)

17 | P a g e


V. Electrically Scanned Difference Patterns The antenna array patterns that have been discussed previously are referred to as sum patterns because they act to transmit and receive in a central window. Sometimes, it is advantageous to attempt the converse- to not transmit and receive in a central window. The antenna pattern which has this feature is referred to as a difference pattern. It is achieved by using a uniform linear array with a 180° phase shift applied to one entire half of the array (figure 12). Please note that in the following treatment, the array is assumed to have 2N elements.

Figure 13: ULA Difference Array Configuration

In order to describe this, the contributions of the array are handled separately, as in above in Section II. The Array factor is now the sum of two array factors:

18 | P a g e


∑

(

)

(

∑

(

∑

)

)

( )

(

∑

)

This can be simplified to:

Where S1 and S2 are given by: ∑

(

)

(

∑

)

Both of the above terms have the form of a geometric series. This means that their sum has the form: (

∑

)

(

(

)

)

Consequently, it can be shown that: ( (

) )

( (

) )

Therefore, the array factor can be shown to be: ( (

) ) (

( ( ( (

)

) )

( (

) )

) )

In order to electrically scan the central minimum of the array pattern, an appropriate inter element phase shift to be determined. In order to define this, let the case of a minimum be considered. In order for a minimum to exist, the numerator of AF must be zero: ( 19 | P a g e

)

Antenna Theory and Design: 2013 Take Home Project (

)

Furthermore, this implies that the argument of sin() must be an integer multiple of π:

From the above, it can be shown that the angle where the central null is formed is given by: (

)

Which also implies: (

)

In order to test this result, the above results are integrated into a simulation. First the simulation’s ability to create the necessary excitation distribution (figure 13) and difference pattern (figure 14) is tested. Next, the simulation’s ability to scan the main null is tested by scanning the null to 120°, as shown in figure 15.

Figure 14: Antenna Excitation Distribution for a 14 Element ULA

20 | P a g e


Figure 15: Difference Pattern for a 14 Element ULA (d = λ/2)

21 | P a g e


Figure 16: Difference Pattern for a 14 Element ULA (d = λ/2) With Central Null Shifted to 120°

Finally, the above results were also simulated for a tapered difference array. Figures 16 and 17 show the distribution and the unshifted pattern, respectively. Notice that no null except the central null in the pattern drops below -21.56dB. Figure 18 shows that the same method used to scan the null of the ULA difference pattern will also shift the null of the tapered distribution.

22 | P a g e


Figure 17: Element Distribution for the 14 Element Tapered Difference Pattern

Figure 18: The 14 Element Tapered Difference Pattern

23 | P a g e


Figure 19: The 14 Element Tapered Difference Pattern (shifted to 120°)

24 | P a g e


VI. Planar Inverted F Antennas The final section of this paper shall consider the Planar Inverted F Antenna (PIFA). As will be shown, the PIFA allows monopole antenna-like performance, but in a much less obtrusive package. The PIFA consists of a ground plane that is shorted to a parallel plate. The parallel plate is then fed coaxially through a hole in the ground plane. For an example, see figure 19. When multi-band performance is required, an “L” shaped slot is often cut in the top plate of the PIFA. In order to study this in greater detail, multiple configurations were studied (see figure 20).

Figure 20: The Planar Inverted F Antenna (PIFA)

Figure 21: Multiband PIFA Configurations Studied

25 | P a g e

Antenna Theory and Design: 2013 Take Home Project The return loss, impedance, and far field antennas of each of multiple configurations was studied in order to determine the effect of increasing/decreasing slot width and moving the right angle bend in the slot further from the feed location. The results of these experiments is tabulated in table 4 and figures 21 through 25.

Figure 22: Performance of the No Slot Case

Figure 23: Performance of the Control Case

26 | P a g e


Figure 24: Performance of the Shifted Slot Case

Figure 25: Performance of the Small Slot Case

27 | P a g e


Figure 26: Performance of the Large Slot Case

As is visible in figures 21-25 above, the addition of a slot adds a second resonance to the PIFA. The width of the slot appears to control how strongly coupled the “second antenna” is. When the slot is wide, the second resonance has a wide bandwidth. The shift in the slot and also the large slot also appear to increase the frequency of the second resonance. The wide slot also appears to increase the frequency of the secondary resonance. This is likely caused by the length of the tab created by the “L” shaped cut. The length of a PIFA’s top plate will determine its resonant frequency- with longer plates causing lower resonant frequencies [3]. In essence, a PIFA with a slot in it acts like two coupled antennas. The shorter antenna is the shorter tab which corresponds to a higher frequency resonance, while the “L” shaped longer tab corresponds to the lower frequency resonance. In order to create a dual band PIFA, the overall size of the PIFA is scaled to realize an appropriate lower frequency. Then, the length of the small tab is adjusted to provide an appropriate higher frequency resonance. In order to achieve a higher frequency resonance, make the tab shorter. In any case, the change in slot configuration has a relatively large impact on the frequency of the secondary resonance- around 30% variation among the cases studied here.

28 | P a g e

Antenna Theory and Design: 2013 Take Home Project F1 F2 RL @ F1 R @F1 X @F1 (GHz) (GHz) (dB) (Ω) (Ω) Effect (relative to control) 1.5 2.08 -26 52.3 4.5 N/A

Case Control

Shift slot

1.56

Large Slot

1.54

Small Slot

1.53

No Slot

2.17

2.55

2.08

1.59 N/A

-50

-35

-31 -28

50.1

Higher 2nd resonant frequency. Primary resonance is relatively -0.3 unaffected.

51.5

Higher 2nd resonant frequency. Primary resonance is relatively 1 unaffected.

50.5

Tighter coupling of secondary resonance. Primary resonance is -2.8 relatively unaffected.

48.8

Primary resonance is relatively -3.7 unaffected with/without the slot.

Table 4: Parametric Study of the Effects of a Slot on the Performance of a PIFA

In order to demonstrate how one may design a dual band PIFA with an appropriate primary and secondary resonance, this paper will design single and dual band PIFAs with primary resonance in the 4G LTE Band 14 frequency band of 758-798MHz. The design goal will be specifically to place the primary resonance in the middle of the band- around 778MHz. Then, a dual band antenna will be designed that implements the same primary resonant behavior but also:  

The highest possible secondary resonant frequency (below 2GHz) The lowest possible secondary resonant frequency

The dimensions of the PIFA will be scaled from a PIFA presented in a paper by Virga and Rahmat-Samii [3]. The overall width and height of the PIFA will be divided by a scale factor denoted in figures as “SF”.

Figure 27: Original PIFA from which all PIFAs in this Paper Are Scaled [3]

29 | P a g e

Antenna Theory and Design: 2013 Take Home Project For the purpose of the following discussion, the modifications to the above PIFA will be referred to by a scale factor SF, which corresponds to the amount each of the above dimensions is divided down, and a scale factor SFt, which corresponds to the location and length of the slot. For more details, refer to Appendix B. The first stage in the design is the adjustment of the overall size of the PIFA to match it to 778MHz. A full wave simulation tool, HFSS, was used to simulate various scale factors in order to determine the optimal value of the overall scale factor SF. The results of these simulations indicate that a value of 0.8642 is appropriate to get a primary resonance of 778MHz (see figure 27). The performance of the PIFA is shown in figure 28.

Figure 28: Scale Factor Search Resulting in 780MHz Primary Operating Frequency

Figure 29: Performance of PIFA for SF = 0.8642

30 | P a g e

Antenna Theory and Design: 2013 Take Home Project The radiation patterns of the PIFA are a fair match for the patterns provided in [3] and are quite broad. The primary resonance is 780MHz and has a good match to 50Ω with an impedance of 49.3+j0.9Ω. The same scale factor is applied to a PIFA with a 3mm wide slot and SFt = 1.5. The primary resonance was slightly affected so the PIFA was rescaled to have SF=0.91 and SFt=1.5. The performance characteristics of the resulting device are summarized in figure 29. The resonant frequencies of the antenna are 770MHz and 1.32GHz. The antenna has a good match to the feed with and impedance of 59+j4.5Ω at 770MHz and an impedance of 43.2+j1.7Ω at 1.32GHz.

Figure 30: First Attempt at a Dual Band PIFA Performance Results

Two different antennas were obtained from this configuration by changing the size of the tab. The size of the tab was decreased in order to increase the frequency of the secondary resonance. The size of the tab was increased in order to decrease the frequency of the secondary resonance. Like the scale factor SF, the appropriate value of the scale factor SFt was selected by a manual search of values based on simulation results. Figure 30 illustrates how the simulation results are used to adjust tab size.

31 | P a g e


Figure 31: Simulation Results are Used to Adjust the Frequency of Second Resonance

The design which maximizes the secondary frequency of resonance utilizes a slot of width 3mm, SF = 0.91, and SFt = 2. This design achieves a primary resonance frequency of 770MHz and a secondary resonance of 1.68GHz. The impedances at these frequencies are 55.4+j4.5Ω and 49+j2.9Ω, respectively. The radiation patterns are broad, although the pattern at the secondary resonance is less well behaved than the pattern at primary resonance.

Figure 32: Performance Characteristics of First Dual Band PIFA Design

32 | P a g e

Antenna Theory and Design: 2013 Take Home Project The design which minimizes the secondary frequency of resonance utilizes a slot of width 3mm, SF = 0.91, and SFt = 1.25. This design achieves a primary resonance frequency of 810MHz and a secondary resonance of 1.15GHz. The impedances at these frequencies are 50.3+j1.6Ω and 50.1+j2.8Ω, respectively. The radiation patterns are broad, although the pattern at the secondary resonance is less well behaved than the pattern at primary resonance.

Figure 33: Performance Characteristics of Second Dual Band PIFA Design

33 | P a g e


Dimension A (mm) B (mm) C (mm) D (mm) E (mm) F (mm) G (mm) H (mm) I (mm)

No Slot

Control

65.32 34.17 N/A 14.52 1.78 2.81 N/A 3.04 N/A

65.32 34.17 11.16 14.52 1.78 2.81 1 3.04 14.52

Build your own dual band PIFA

SF SFt

Large Slot Small Slot Shift Slot 65.32 34.17 11.16 14.52 1.78 2.81 3 3.04 14.52

65.32 34.17 11.16 14.52 1.78 2.81 0.1 3.04 14.52

65.32 34.17 12.15 14.52 1.78 2.81 1 3.04 14.52

1 1

Table 5: PIFA Dimensions, Including a PIFA Dimension Calculator

Figure 34: Dual Band PIFA Dimension Definition

34 | P a g e

780 MHz 770 770 810 Single MHz/1.32 MHz/1.68 MHz/1.15 Band PIFA GHz PIFA GHz PIFA GHz PIFA 134.38 127.62 127.62 127.62 70.30 66.76 66.76 66.76 N/A 20.25 15.19 24.30 29.87 28.36 28.36 28.36 0.86 0.91 0.91 0.91 5.79 5.49 5.49 5.49 N/A 3 3 3 6.25 5.94 5.94 5.94 N/A 28.36 28.36 28.36

Build Your Own Dual Band PIFA 116.13 60.75 30.37 25.81 1.00 5.00 3 5.41 25.81


35 | P a g e


IX. Conclusion This paper examined several topics utilizing simulation techniques in antenna theory. The first was that of tapered and uniform linear arrays. A simulation was devised that enabled study of the properties of the uniform linear antenna array and tapered array in both sum and difference pattern configurations. Differences between the approximate theoretical and simulation were discovered and explained as a failing in the approximate theory due to approximations. The closed form solution was used to demonstrate this. A method of design a tapered array by simulation via a linear taper was presented and performance data for this configuration was presented. The linear taper was applied to study a difference pattern and steerable array patterns. Next, the use of HFSS to analyze a PIFA was demonstrated. HFSS was used to study changes to the PIFA that enable dual band performance. Several PIFA designs were implemented that provide performance sufficient for operation in the 4GLTE band 14 frequency range. Several design guidelines were noted and demonstrated.

36 | P a g e


X. References [1] K. Melda, “ECE 584 Antenna Theory and Design, Spring 2013: Take Home Midterm Project”. Unpublished work. Tucson AZ, March 2013 [2] C. Balanis, “Antenna Theory: Analysis and Design”. 3rd ed. John Wiley and Sons, Inc. Hoboken, NJ 2005 [3] K. Virga, Y. Rahmat-Samii, “Low-Proﬁle Enhanced-Bandwidth PIFA Antennas for Wireless Communications Packaging” in IEEE Transactions on Microwave Theory and Techniques, Vol. 45, No. 10, Oct 1997

37 | P a g e


XI. Appendix A: MATLAB Simulation of A Linear Antenna Array %-------------------------------------------------------------------------% %Housekeeping- clean up previously used variables and the command screen %-------------------------------------------------------------------------% clear; clc; %-------------------------------------------------------------------------% %Initialize Variables and Constants %-------------------------------------------------------------------------% %Basic variables lambda = 1; k = 2*pi / lambda; radians = pi/180; degrees = 1; %Basic System/Design Parameters d = 1/2 * lambda; M = input('Please enter number of elements-\n>'); NoiseFloor = 1e-6; Excite_Mode = input('\nPlease select mode of excitation- \n1 = … Uniform\n2 = Difference Pattern\n3 = Tapered Sum\n4 = … Tapered Difference\n>', 's'); switch Excite_Mode case '1', %Uniformly excited amplitude sum pattern Excitation = ones(1,M);%No taper, no phase shift between sides of array case '2', %Uniformly excited amplitude difference pattern Excitation = [ones(1,M/2),-ones(1,M/2)];%No taper, pi phase between sides case '3', %A linear taper between 1 at center of array and 0.1 at edge of %array is used to synthesize a tapered sum pattern a = 0.5;%input('Please enter taper parameter-\n>'); Excitation = ([linspace(a,1,M/2),linspace(1,a,M/2)]); case '4' %A linear taper between 1 at center of array and 0.1 at edge of %array is used to synthesize a tapered difference pattern a = 0.5;%input('Please enter taper parameter-\n>'); Excitation = ([linspace(a,1,M/2),-linspace(1,a,M/2)]); otherwise, disp('Error: invalid excitation mode selected- Aborting simulation.') return end SteeringAngle = input('\nPlease select steering angle (in degrees)-\n>'); %Initialize working variables Theta = (0:.01:180) * degrees; Beta = -k * d * cos(SteeringAngle*radians);

38 | P a g e

Antenna Theory and Design: 2013 Take Home Project Psi = k*d*cos(Theta*radians) + Beta; Efield = 0; %-------------------------------------------------------------------------% %Calculate the electric field due to superposition of antenna elements %-------------------------------------------------------------------------% for element = (1:1:M) Efield = Efield + Excitation(element)*exp(1i.*(element-1).*Psi); end %-------------------------------------------------------------------------% %Normalize the electric field due to the M different antenna elements %-------------------------------------------------------------------------% Efield = Efield / M; %-------------------------------------------------------------------------% %Calculate the Amplitude due to the electric field %-------------------------------------------------------------------------% Amplitude = abs(Efield); Amplitude = 20 * log10(Amplitude + NoiseFloor); %-------------------------------------------------------------------------% %Normalize the total Amplitude to 0dB as max %-------------------------------------------------------------------------% Amplitude = Amplitude - max(Amplitude); %-------------------------------------------------------------------------% %Plot angular amplitude distribution %-------------------------------------------------------------------------% figure(1); plot(Theta, Amplitude); %-------------------------------------------------------------------------% %Find 3dB Points %-------------------------------------------------------------------------% [Y,Imax]=max(Amplitude+3); [Y,Izero]=min(abs(Amplitude+3)); FWHM = 2*abs(Theta(Imax)-Theta(Izero)); disp(' '); disp([' The FWHM is ', num2str(FWHM,5),'.']); disp(' '); %-------------------------------------------------------------------------% %Find Local Maxima %Tabulate Results %-------------------------------------------------------------------------% [pks,locs] = findpeaks(Amplitude); ThetaMaxima = Theta(locs)'; disp(' Theta Amplitude') disp([Theta(locs).' Amplitude(locs).']) %-------------------------------------------------------------------------% %Calculate Directivity %-------------------------------------------------------------------------% P_Rad = trapz(Theta*radians,sin(Theta*radians).*10.^(Amplitude/10))*2*pi;

39 | P a g e

Antenna Theory and Design: 2013 Take Home Project U_Max = max(10.^(Amplitude/10)); Directivity = 4*pi*U_Max/P_Rad; disp(' '); disp([' The directivity is ', num2str(Directivity,5), '.']);

40 | P a g e

Antenna Theory and Design Project

Recommend Documents