Accurate measurement of the input impedance of bow-tie antenna by the S-parameter method

Transcription

1 Accurate measurement of the input impedance of bow-tie antenna by the S-parameter method Kazuma Endo 1a), Takayuki Sasamori 2, Teruo Tobana 2, and Yoji Isota 2 1 Graduate School of Systems Science and Technology, Akita Prefectural University, 84 4 Tsuchiya-Ebinokuchi, Yurihonjo, Akita , Japan 2 Faculty of Systems Science and Technology, Akita Prefectural University, 84 4 Tsuchiya-Ebinokuchi, Yurihonjo, Akita , Japan a) M17B005@akita-pu.ac.jp Abstract: Recently, the S-parameter method, which is the measurement method of the input impedance of the balanced antenna, has been proposed. In this paper, to show the possibility of the S-parameter method using jig fabricated microstrip line, the input impedance of a bow-tie antenna is examined. As a result, it is observed that the results of the S-parameter method are shown to agree with the other results at frequencies less than 3 GHz. In addition, it is thought that the main cause of the measurement error of the S-parameter method using the MSL type jig is due to the electromagnetic coupling between the antenna element and MSL. Keywords: S-parameter method, balanced fed antenna, bow-tie antenna, input impedance, modified open correction Classification: Antennas and Propagation References [1] K. Sato, K. Nishikawa, N. Suzuki, and A. Ogawa, Analysis of antennas mounted on portable equipment near human body, IEICE Trans. Commun. (Japanese Edition), vol. J79-B-II, no. 11, pp , Nov [2] H. Morishita, Y. Kim, and K. Fujimoto, Design concept of antenna for small mobile terminals and the future perspective, IEEE Antennas Propag. Mag., vol. 44, no. 5, pp , Oct DOI: /MAP [3] K. Fujii and T. Iwasaki, Evaluation of site for measuring complex antenna factors: comparison of theoretical calculation and TRL-based experiment, IEICE Trans. Commun., vol. E83-B, no. 10, pp , Oct [4] R. Meys and F. Janssens, Measuring the impedance of balanced antennas by an S-parameter method, IEEE Antennas Propag. Mag., vol. 40, no. 6, pp , Dec DOI: / [5] T. Sasamori and T. Fukasawa, S-parameter method and its application for antenna measurements, IEICE Trans. Commun., vol. E97-B, no. 10, pp , Oct DOI: /transcom.E97.B.2011 [6] T. Sasamori, K. Endo, T. Tobana, and Y. Isota, A Study of Jig for S-parameter 158

2 method using microstrip line Part 5, IEICE Technical Report, AP , pp , Apr Introduction It is known that the antenna characteristics of the portable radio terminal are changed greatly when the body of the terminal is held in the hand [1]. This problem is caused by the fact that the current distribution on the terminal changes by the hand. The effect of a balanced fed antenna was reported to reduce the influence of holding the terminal by hand [2]. The input impedance of the balanced antenna is measured conventionally by using a balun that forces opposite currents into each part of the radiation elements. To eliminate the characteristics of the balun from the measurement results, the measurement method using the TRL calibration is reported [3]. However, since the procedure of this measurement method is complicated, there is a disadvantage that increasing uncertainty. Recently, to measure the impedance of the balanced antenna, the S-parameter method has been developed that uses a jig instead of a balun, and the two-port vector network analyzer (VNA) [4]. Because no balun is used, it is possible to measure impedance over a wide frequency bandwidth. However, when the measured frequency rises, measurement accuracy decreases because the influence of the jig cannot be disregarded. We have proposed the modified open correction using the ABCD-matrix that can remove the influence of the jig including an open end impedance of the transmission line, and reported on the calculated result of an input impedance of a dipole antenna using a jig fabricated by coaxial cable [5]. On the other hand, only few attempts have so far been made at confirmation of the effect of the modified open correction for a jig made of microstrip line (MSL type jig) that can be fabricated with the machining or etching [6]. And, an impedance measurement of the wideband antenna such as a bow-tie antenna by the S- parameter method using the MSL type jig (MSL S-parameter method) has never been examined. In this paper, to show the possibility of the MSL S-parameter method, the input impedance of a bow-tie antenna, which is one of the broadband balanced antennas, is examined. Measured and calculated results of the MSL S-parameter method are compared with a calculated result of a bow-tie antenna fed by a delta gap source and a measured result of a bow-tie monopole antenna on a ground plane. Additionally, influence by the direction of antenna element on the MSL is examined. 2 S-parameter method and antenna structure Fig. 1(a) shows a measurement system for input impedance of a bow-tie antenna by the MSL S-parameter method. The measurement jig is composed of two dielectric substrates called jigs 1 and 2. The MSL is printed on the dielectric substrate, and an SMA connector is soldered to one end of the MSL. The two dielectric substrates are fixed back-to-back with nylon screws, as shown in Fig. 1(b). Elements of the bowtie antenna are soldered to the tips of MSL of the jig. After the VNA is calibrated at 159

3 the end of coaxial cable using the full two port calibration (short, open, load and through) technique, the S-parameters of the antenna with the jig and the open end jig without antenna element are measured by the VNA. We use the modified open correction to remove the influence of the jig with the open end impedance of the MSL from the result of a measurement [4]. (a) Measurement system. (b) Bow-tie elements with measurement jig. Fig. 1. MSL S-parameter method and bow-tie antenna. 3 Results and discussion Fig. 2(a) shows a calculation model of the MSL S-parameter method. Two antenna elements are isosceles trapezoid whose height is 34.8 mm, length of upper base is 71.0 mm, length of lower base is 1.6 mm and thickness is 0.4 mm. The MSL type jig is composed of two dielectric substrates with dimensions of :8 mm 3 and relative permittivity of The MSL with a length of 20 mm, width of 2.4 mm, and characteristic impedance of 50 Ω is printed on the dielectric substrate. In this section, the antenna elements are installed perpendicularly to the MSL. When the antenna element is not installed to the MSL, the open end impedance can be shown by capacitance as C ¼ 1=ðj!Z L Þ¼0:035 pf, which is calculated from the electric and magnetic fields at the end of the MSL using the FDTD method. Fig. 2(b) shows a calculation model of bow-tie antenna fed by a delta gap source. Two radiation elements are isosceles right triangles whose base length is 71.0 mm, height is 35.6 mm and thickness is 0.4 mm. The elements are fed between their vertex angles. Fig. 2(c) shows a photo of bow-tie monopole antenna on the ground plane with the dimensions of 1:25 1:25 m 2. Doubled measured input impedance of the antenna on the ground plane is compared with that of bow-tie antenna. Figs. 2(d) and (e) show the frequency characteristics of the input impedance of the bow-tie antenna. The results of the MSL S-parameter method are corrected using the modified open correction. Red line and red plots show calculated and measured results of the MSL S-parameter method, respectively. As can be seen, they are in good agreement. It is thought that the cause of a little difference between 160

4 the calculated and measured results is fabrication error. On the other hand, the calculated result of the delta-gap source model agrees with the measured result of the bow-tie monopole antenna very well. Regarding the resistance shown in Fig. 2(d), it is observed the results of the MSL S-parameter method and the other results are in a good agreement over the wide frequencies. As for the result of the reactance shown in Fig. 2(e), it is notable that the difference between the MSL S-parameter method and other results is comparatively large in the frequency above 3 GHz. In order to examine the influence of the jig, we calculated the MSL S-parameter method using a jig whose dimensions are 0 0 1:6 mm 3. Consequently, the results of the MSL S-parameter method fairly agree with the calculation result of the delta gap source model. Therefore, it is considered that the cause for the difference of reactance is the electromagnetic coupling of the antenna element and the jig. (a) Calculation model of MSL S-parameter method. (b) Calculation model fed by delta gap source. (c) Bow-tie monopole antenna on ground plane. (d) Resistance (e) Reactance Fig. 2. Calculated and measured input impedance of bow-tie antenna. Then, we investigate the influence by the direction of antenna element on the MSL. One configuration is the case where the antenna elements are installed perpendicularly to the MSL as shown in Fig. 3(a). In this case, we call the model type A, and this is the same as Fig. 2(a). And, another configuration where the antenna elements are installed in parallel to the MSL as shown in Fig. 3(b) is called type B. Figs. 3(c) and (d) show the calculated results of the input impedance of the bow-tie antenna. In both figures, it can be seen that results of Type A are nearer to the calculated results of the delta gap source model than Type B. Because the 161

5 distance between the MSL of the jig and the bow-tie antenna element of type B is smaller than type A, it is thought that stronger electromagnetic coupling occurred, and a bigger difference appears in the result of the MSL S-parameter method. (a) Type A (perpendicular model). (b) Type B (parallel model). (c) Resistance (d) Reactance Fig. 3. Comparison of input impedance of bow-tie antenna whose mounting directions are different. 4 Conclusion In this paper, to show the possibility of the MSL S-parameter method, the input impedance of a bow-tie antenna has been examined. First, we explained the measurement system of the MSL S-parameter method and the bow-tie antenna elements with the measurement jig. Next, measured and calculated results of the MSL S-parameter method using the jig with dimensions of :6 mm 3 were compared with the measured result of the bow-tie monopole antenna on the ground plane and the calculated result of delta gap source model. Furthermore, we examined the influence of configuration on the result of the MSL S-parameter method. The results of the input impedance of the bow-tie antenna by the MSL S-parameter method were shown to agree with other results at frequencies less than 3 GHz. It is thought that the main cause of the measurement error of the MSL S-parameter method is due to the electromagnetic coupling between the antenna element and MSL on the jig. 162