A Yagi-Uda antenna made of high-permittivity ceramic material PDF

Preview

Summary

2013 Loughborough Antennas & Propagation Conference 11-12 November 2013, Loughborough, UK A Yagi-Uda Antenna Made of High-Permittivity Ceramic Material Jan Hesselbarth, Daniel Geier Miguel Angel Bueno Diez University of Stuttgart Universität der Bundeswehr München Institute of Radio Frequency Te...

Description

2013 Loughborough Antennas & Propagation Conference

11-12 November 2013, Loughborough, UK

A Yagi-Uda Antenna Made of High-Permittivity Ceramic Material Jan Hesselbarth, Daniel Geier

Miguel Angel Bueno Diez

University of Stuttgart Institute of Radio Frequency Technology 70550 Stuttgart, Germany

Universität der Bundeswehr München Chair in High-Frequency Technology and Mobile Communication, 85577 Neubiberg, Germany

Dielectric resonator antennas (DRA) in the RF domain are known for compact, mechanically simple and stable structure, combined with high efficiency. Comparing DRA to metal-wire Yagi antennas, however, the latter has superior efficiency, probably lower weight and shows mechanical simplicity, too. Dielectric Yagi antennas might therefore find use only in niche applications, for example, in environments which are chemically unfavorable to metals.

Abstract—Half-wavelength dipole and 3-element Yagi antennas made of high-permittivity dielectric ceramic rods instead of metal wires are described. For a frequency of operation around 3.7 GHz, dielectric rods of 4x4 mm2 crosssection and relative permittivity of 34 can replace the metal wires in these antennas. A structure allowing to feed the dielectric dipole by means of a balanced (two-wire) transmission line is described. The 3-element dielectric Yagi antenna shows a simulated directivity of 8.0 dBi, a simulated as well as measured front-to-back ratio of 10.2 dB, a measured respectively simulated -10 dB impedance bandwidth of 11.6% and 9.9% at 3.7 GHz, and a simulated radiation efficiency of 94%.

The following text describes simulation, built and measurement of dielectric Yagi antennas. Furthermore, in order to connect the antenna to co-axial cable, a balanced feed structure and a tapered-line balun are needed and will be described in detail.

Keywords— Yagi antennas, dielectric resonator antennas

I.

INTRODUCTION

The construction of a directive antenna consisting of a fed dipole, a close-by metal reflector rod and a number of radiation-coupled director rods has been known as Yagi-Uda antenna for a long time [1][2]. These antennas show an advantageous combination of features, such as low weight, high efficiency, high gain and relatively large impedance bandwidth. More recently, the principle of the Yagi antenna found its way into the infrared frequency range, by applying primarily gold [3] or silver [4] material rods, noting that properties of metals at optical frequencies are well different from their radio-frequency (RF) behavior.

II.

DESIGN OF THE DIELECTRIC DIPOLE

The dielectric Yagi antenna is built using high-permittivity ceramic rods. The permittivity is εrel = 34 and the loss tangent is roughly tan δ = 0.0001 at the frequency of operation. Simulations indicate that such a small loss tangent has practical no influence on antenna efficiency. The cross-section of the ceramic rods is a square of 4 mm x 4 mm. The design process starts with the driven dipole element. As it is known for conventional metal dipoles of “thick” cylindrical cross-section, the feed impedance has a large reactive part. In addition, the feed impedance is strongly dependent on the geometry of the feeding gap region, such as gap distance and dipole cross-section [8][9]. For the dielectric rod dipole, it turns out that an additional difficulty is the transition from the metallic feed-line (which carries the conduction currents) to the dielectric (which is supposed to carry the displacement currents). This transition should not become too localized. Furthermore, direct contact (without any air gap) between the metallic feed-line and the dielectric rods is advantageous but difficult to realize in practice unless the ceramic is plated with metal. For these two reasons, we designed a metal clamp structure which is soldered to the metallic (balanced, two-strip) feed-line. On the other end, the clamp embraces the dielectric rods but keeps a small air-gap between ceramic and metal with the help of a plastic (PMMA) splint. The air-gap (about 0.1 mm) makes the structure less

In the RF domain, dielectric resonator antennas made of high-permittivity brick, rod, or puck radiators are well-known (see, e.g., [5] and references therein). Recently, bricks of particularly large aspect ratio have been employed as antennas [6][7]. Their resonance mode named TExδ01 shows the electric field primarily oriented in the direction of the long axis of the brick, and concentrated in the high-permittivity dielectric (εrel = 90). In the following, we will use high-permittivity dielectric rods to replace the metal rods of a conventional Yagi antenna. This is straightforward considering the fact that by doing so, conduction current along the metal rods is simply replaced by displacement current in the dielectric rods. Nevertheless, some optimization is necessary because the permittivity of practically available material (in our case, εrel = 34) is limited.

978-1-4799-0091-6/13/$31.00 ©2013 IEEE

94

sensitive to mechanical tolerances. The metal clamp has a length of 13.34 mm in contrast to the overall extension of the ceramic dipole of 46.35 mm, therefore, the metal clamp is not contributing to the radiation. For reasons of mechanical stability, the feeding gap between the high-permittivity ceramic rods is filled with the circuit board dielectric material of the feed-line. Fig. 1 shows the details of the feeding region of the dipole. As an additional optimization parameter and in order to allow the plastic (PMMA) splint maintain mechanical stability, cutouts in the ceramic rods reduce their cross-section to 4 mm x 1.7 mm in the feeding region. This feature also improves the omnidirectional characteristics of the dipole radiation pattern. Taking into account the constraints of the mechanical structure and the available materials, the feed impedance of the balanced two-strip feed-line was optimized using electromagnetic CAD software (finite element simulation) to about (44 + j0) Ω at 3.8 GHz. Using a roughly 100 mm long tapered transition, the 44 Ω balanced two-strip line is converted in a 50 Ω microstrip line with a SMA connector attached. Fig. 2 shows a plot of the simulated electric field of the dipole at 3.8 GHz.

III.

DESIGN OF THE DIELECTRIC YAGI ANTENNA

The dielectric driven dipole from the previous section is now combined with a dielectric reflector rod oriented in parallel to the driven element. Separation distance between the two elements and length of the reflector rod were optimized with respect to maximum directivity and maximum front-toback ratio. A distance between driven dipole and reflector of about a quarter wavelength is advantageous, which is in agreement with conventional (metal) Yagi antenna designs [10]. The length of the dielectric rod reflector, however, turns out to be “the longer the better” (at least, as shown in simulations, up to a length of two wavelengths), and was limited for practical reasons to about one free-space wavelength. Directivity and front-to-back ratio cannot be maximized both at the same time. The simulation of the final two-element antenna (including balun and feed-line) shows an optimum directivity of 5.8 dBi and a front-to-back ratio of 4.0 dB at 3.6 GHz. Fig. 3 shows a picture of this two-element dielectric antenna. Low-permittivity low-loss closed-cell foam (Rohacell; εrel = 1.06) was added in order to improve mechanical stability without disturbing the electromagnetic fields.

Fig. 1. Dielectric dipole with feed structure and feed-line taper. All dimensions in millimeters. There is an air-gap of 0.1 mm between the clamp and the ceramic rod.

Secondly, the dielectric driven dipole is combined with a dielectric director rod oriented in parallel to the driven element. Separation distance between the two elements and length of the director rod are optimized with respect to maximum directivity and maximum front-to-back ratio. It turns out that for increased front-to-back ratio, the separation distance between the two ceramic rods reduces to less than 0.1 λ and in those cases it will deteriorate the feed impedance significantly. This is considered as not acceptable. Directivity and front-to-back ratio cannot be maximized both at the same time. The

simulation of the final two-element antenna (including balun and feed-line) shows an optimum directivity of 6.0 dBi and a front-to-back ratio of 9.2 dB at 3.8 GHz. The 3-element dielectric Yagi antenna is simply the combination of the two, previously optimized two-element structures described above. That is, the geometrical dimensions of the 3-element dielectric Yagi antenna are those of the respective two-element structures. Note that even for the twoelement structures, the driven dipole remains identical to the

95

E [V/m]: 5000

500

50

Fig. 2. Magnitude and vector plot of the electric field (snapshot in time) of the dielectric dipole with feed structure at 3.8 GHz (note: vertical magnetic symmetry plane applied). Fig. 4. 3-element dielectric Yagi antenna. The lengths of reflector, dipole and director rods are, respectively, 77.6 mm, 46.35 mm, 44.4 mm. All rods have a cross section of 4 mm x 4 mm. The distance (side-to-side) between reflector and dipole is 23.0 mm. The distance (side-to-side) between dipole and director is 10.7 mm.

E [V/m]: 5000

500

50

Fig. 3. Two-element dielectric antenna consisting of dipole and reflector.

previously optimized single dielectric dipole with feed transition. It is known from conventional Yagi antennas [11], that due to mutual coupling effects, such a simple combinatory approach is sub-optimum. For example, the driven element and its feed impedance will definitely vary when reflector or director or both are added. However, due to the electromagnetic complexity of the 3-element dielectric Yagi, a brute force optimization is considered impossible. Nevertheless, a number of parameter variations in the simulated antenna model show that no significant improvement of directivity, front-to-back ratio, impedance match and bandwidth together can be easily obtained. Thus, the 3-element antenna is considered to perform at least not far from optimum.

Fig. 5. Magnitude and vector plot of the electric field (snapshot in time) of the dielectric 3-element Yagi antenna with feed structure at 3.7 GHz.

to-back ratio of 10.2 dB at 3.7 GHz. The simulated radiation efficiency is 94%. Fig. 5 shows a plot of the simulated electric field of the 3-element antenna at 3.7 GHz. Close inspection of the electric field inside the high-permittivity dielectric, which is proportional to the displacement currents, over time or phase reveals that fields in dipole and reflector are almost out-ofphase, that is, they are oriented in opposite directions at almost all time. Fields in dipole and director, however, show a phase difference of very roughly 90°, as they point in the same direction over about half a time period, and are oriented in

Fig. 4 shows a picture of the realized 3-element dielectric Yagi antenna. Low-permittivity low-loss closed-cell foam (Rohacell; εrel = 1.06) was added in order to improve mechanical stability. Simulation of the 3-element Yagi as obtained by combining the two-element structures (including balun and feed-line) shows a directivity of 8.0 dBi and a front-

96

opposite directions over the other half period of time. This observation is consistent with conventional metal-wire Yagi antennas.

IV.

MEASUREMENT OF THE DIELECTRIC YAGI ANTENNA

Measurement of input impedance match is reported in Fig. 6 and shows good agreement with simulation. The measured resonance frequency dip differs from the simulated one by 0.07 GHz. The measured -10 dB bandwidth is 11.6% versus 9.9% for simulation. Measurements of the far-field radiation pattern of the 3element dielectric Yagi antenna were done using a two-axis spherical measurement system in an anechoic chamber at different frequencies. Fig. 7 shows the definition of cut-planes and angles for the measurement of horizontal-cut co- and cross-polarization pattern, reported in Fig. 8, and vertical-cut co- and cross-polarization pattern, reported in Fig. 9.

Fig. 6. Input match of the 3-element dielectric Yagi antenna. Solid red curve: measurements. Dashed blue curve: simulation.

The co-polarization pattern show the front-to-back ratio of about 10 dB as expected from the simulations. The vertical-cut co-polarization pattern shows an asymmetric behavior due to the asymmetric feed structure (clamp, splint etc.) of the dipole (not due the balun and feed-line as verified by simulations). The cross-polarization pattern show low cross-polarization below -20 dB in the horizontal-cut and even below -30 dB in the vertical cut. The “dip” around 180° in the horizontal-cut cross-polarization is a measurement artifact: in the two-axis spherical measurement system, the mounting beam for the antenna points into this direction, thereby shadowing any radiation. This sector is hatched in the plots.

V.

CONCLUSION

A half-wavelength dipole antenna and a 3-element Yagi antenna made of high-permittivity dielectric ceramic rods instead of metal wires have been presented for operation at 3.7 GHz. The use of dielectric rods with 4 mm x 4 mm crosssection and relative permittivity of 34 was demonstrated and shows that ceramic material can replace the metal wires in these antennas achieving good operation in terms of directivity, front-to-back ratio, impedance bandwidth, and radiation efficiency. The measured and simulated results agree well.

Fig. 7. Definition of cut-planes and angles for the measured far-field pattern. [5]

A. Petosa, A. Ittipiboon, “Dielectric resonator antennas: a historical review and the current state of the art,” IEEE Antennas Prop. Mag., vol. 52, October 2010, pp. 91-116. [6] Y. Gao, Z. Feng, L. Zhang, “Experimental investigation of new radiating mode in rectangular hybrid dielectric resonator antenna,” IEEE Ant. Wireless Prop. Lett., vol. 10, 2011, pp. 91-94. [7] Y. Gao, Z. Feng, L. Zhang, “Investigation of a new radiating mode and the traditional dominant mode in rectangular dielectric resonator antenna,” IEEE Ant. Wireless Prop. Lett., vol. 11, 2012, pp. 909-912. [8] R.W.P. King, Tables of Antenna Characteristics, IFI/Plenum Data Corp., New York, 1971. [9] D.H. Werner, “A method of moments approach for the efficient and accurate modeling of moderately thick cylindrical wire antennas,” IEEE Trans. Antennas Prop., vol. 46, no. 3, March 1998, pp. 373-382. [10] P.P. Viezbicke, “Yagi antenna design,” NBS Technical Note, no. 688, December 1976. [11] R.W.P. King, G.J. Fikioris, R.B. Mack, Cylindrical Antennas and Arrays, Cambridge University Press, Cambridge / UK, 2002.

REFERENCES [1] [2]

[3]

[4]

H. Yagi, “Beam transmission of ultra short waves,” Proc. IRE, vol. 16, June 1928, pp. 715-741. D.M. Pozar, “Beam transmission of ultra-short waves: An introduction to the classic paper by H. Yagi,” Proc. IEEE, vol. 85, November 1997, pp. 1857-1863. P. Mühlschlegel, H.-J. Eisler, O.J.F. Martin, B. Hecht, D.W. Pohl, “Resonant optical antennas,” Science, vol. 308, 10 June 2005, pp. 16071609. J. Li, A. Salandrino, N. Engheta, “Shaping the beam of light in nanometer scales: a Yagi-Uda nanoantenna in optical domain,” www.arxiv.org/abs/cond-mat/0703086.

97

Fig. 9. Measured vertical-cut co-polarization (top) and cross-polarization (bottom) pattern.

Fig. 8. Measured horizontal-cut co-polarization (top) and cross-polarization (bottom) pattern. In a sector between 150° and 210° (marked with hatches), results are invalid due to the measurement setup (antenna mounting beam in direction of 180°).

98...