диафрагмированные волноводные фильтры / d1737df3-3ad5-4974-b075-4863c40d7afe

MSMW'13, Kharkov, Ukraine, June 23-28

EXPERIENCE IN DEVELOPING Ka-BAND WAVEGUIDE FILTER WITH

HTS E-PLANE INSERT

V.N. Skresanov1), A.A. Barannik1), N.T. Cherpak1), Y. He2), V.V. Glamazdin1), V.A. Zolotaryov1),

A.I. Shubny1), L. Sun2), J.Wang2), Y.Wu3)

1)O.Ya. Usikov Institute for Radiophysics and Electronics, National Academy of Sciences of Ukraine, 12, Acad. Proskura Street, Kharkiv 61085, Ukraine, e-mail: valery.skresanov@gmail.com

2)Institute of Physics, Chinese Academy of Sciences, Beijing 100190, China, e-mail: yshe@aphy.iphy.ac.cn 3)University of Science and Technology, 100083 Beijing, China

To improve the performance of receivers of both satellite microwave links and radio telescopes, developing the bandpass filter (BPF) with low insertion loss is the actual task. The filter is installed at LNA input and jointly cooled to cryogenic temperatures. Every 0.1 dB insertion loss in BPF increases the receiving system noise temperature for 7-7.5 K at 300 K, which should be considered as notable addition, if we consider that noise temperature of the modern transistor amplifiers in the Ka range is about 60 K when cooled to the liquid nitrogen temperature. The mentioned noise addition depends on operating temperature and level of insertion loss. For a correctly designed multy-pole BPF, level of insertion loss in the passband is ultimately determined by the unloaded Q-factor of individual resonators. The introduction of HTS elements in the filter resonators increases the unloaded Q, therefore, should reduce insertion loss. The idea of the application of HTS materials in the waveguide E-plane BPF was expressed earlier, e.g., in [1], where the authors proposed to cover the waveguide internal walls with the bulk HTS material or to replace the normal metal E-plane insert by HTS one. This idea was advanced in [2] while creating two-pole X-band filter with HTS E-insert. However, the promising directions of developing HTS E-plane filters remained unclear.

This paper presents the results of the design and testing of seven-pole BPF Ka band with bilateral HTS insert in the E-plane rectangular waveguide. Study of the possibility of reducing the insertion loss in the E-plane filters with HTS inserts compared with losses in the filters with normal metal inserts is a purpose of the work. Fig. 1 shows CST project of BPF, and Fig.2 displays photos of BPF prototype and HTS insert. HTS insert is a single crystal MgO substrate of 0.5 mm thickness, of which the two surfaces are coated with superconducting YBCO films. By photolithography method, identical insert portions of the superconducting layers were removed on both sides of a substrate and thereby "window” chains were formed with a height equal to the height of the standard rectangular waveguide of 7.2x3.4 mm2 section. The lengths of all "windows" and the distance between them along the axis 0Z were calculated as a result of designing the filter. Therefore, individual resonators of seven-pole BPF are rectangular waveguide sections with dielectric MgO plates in the E-plane. The mutual coupling between the resonators and the coupling of the first and the last resonators with matched waveguide are carried out by means of pairs of the below-cutoff waveguide sections formed by the original walls of the waveguide and the conducting YBCO layers. It is clear that the principle of operation of the filter with HTS insert analogous to the principle of BPF widely used in practice with normal metal E-plane inserts [3].

Methods of designing E-plane filters with a dielectric insert and bilaterally deposited metal layers are known when the calculation and optimization of dimensions of the resonators and the coupling elements are based on the solution of the boundary value problem by analytical methods of computational electrodynamics [4,5]. We applied a grid method to calculate the electrodynamic analysis of this structure in the synthesis of insert geometry, using a program package 'CST Microwave Studio'. Our approach to the design of BPF is

MSMW'13, Kharkov, Ukraine, June 23-28

especially convenient if there is a need for the synthesis of BPF, in which some elements are difficult to rigorous analysis.

On the basis of the specification for designing BPF, elements of the prototype filter were calculated by the classical methods of circuit theory [6], as well as the characteristic parameters of BPF, namely, resonant frequencies, the coefficients of mutual coupling and external Q-factor of the BPF resonators. Further the geometric dimensions of pairs of chains of coupled resonators in BPF were synthesized so that the characteristic parameters of the real BPF resonators were equal to the calculated. For this purpose amplitude-frequency characteristics of each pair of the resonators of real BPF were calculated using 'CST Microwave Studio'. The characteristic parameters of pairs of BPF real resonators were restored from the obtained amplitude-frequency characteristics of the pairs in accordance with the technique developed by us earlier [7]. In case of difference between the calculated and restored characteristic parameters, the geometric dimensions of certain elements of coupled resonators changed and the calculation is repeated again. In conclusion, it was enough to make "assembly" of the seven-pole BPF from synthesized fragments (pairs of the resonators) in order to obtain the required BPF on the design task performance.

Fig. 3 shows the results of CST calculation of the characteristics of the synthesized BPF with E-plane HTS insert and, for comparison, the result of measuring the S21 parameter of the manufactured filter. We observe a good agreement of the results of design and measurements.

Fig. 4 shows S21 measurements results for the manufactured BPF (Fig. 2) in the temperature range from 94.7 K to 66.1 K. Measurements were carried out in the technological cryostat cooled by liquid nitrogen, with the possibility of vapor pumping to achieve a temperature below 77K. The frequency responses in Fig. 3 and Fig. 4 were obtained using the vector analyzer N5230A by subtracting the response of input and output waveguides from the amplitude-frequency characteristics of the BPF with sections of input and output waveguides.

Central frequencies within the filter passband and insertion loss at these frequencies were calculated from the measurement data in Fig. 4 for each temperature. The results are presented in Fig. 5 in the form of the temperature dependence of these parameters. The center frequency generally decreases with increasing temperature, due to the change in the geometric dimensions and MgO permittivity. Apparently, the nonmonotonic dependence and a different steepness of the characteristic in different parts of the temperature interval are conditioned by the temperature dependence of the reactance of the HTS film, which may have an influence on the resonant frequency of BPF resonator. Therefore, when calculating the correct center frequency of BPF bandwidth, it is necessary to take in account the mentioned temperature effects. The temperature dependence of the insertion loss explained by the change of the conductivity of HTS films at the transition to the superconducting state. In particular, a drastic change in insertion loss is observed at the critical temperature Tc=88K. At a temperature of 66.1 K the insertion loss 0.2 dB were found to be practically coincided with the CST calculations, which used the following material parameters: the conductivity of cooled waveguide cover

σAg=3.9·108S/m, the conductivity of YBCO σYBCO=1.0·1010S/m and the loss tangent of cooled MgO tanδ=6.0·10-6. The confidence interval of the measured insertion loss is about 0.15 dB, since the losses are

662

MSMW'13, Kharkov, Ukraine, June 23-28

Figure 5. Temperature dependencies of BPF insertion loss and central frequency obtained from the measured frequency responses (see Fig. 4)

calculated as the difference between two large quantities. Calculated amount of insertion loss are also obtained with some error because parameters of the materials were used from the literature.BPF insertion loss depends on the relationship of unloaded Q and external losses in the filter resonators [6]. As it is known, the external losses determine also BPF bandwidth. As a result of the increase of the band at the same unloaded Q losses, the resonators insertion loss will decrease. In addition, the insertion loss of BPF increases with the number of resonators, which is chosen from the requirement to provide a given steepness of BPF slopes [6]. For example, the transition from two-pole filter to seven-pole one increases insertion loss by 3.2.

With the known unloaded Q and external Q-factors of BPF resonators, two-pole filter insertion loss can be estimated from the analytical formula given in [6]. So for the filter with the 500 MHz bandwidth external Q- factor is 80. If the unloaded Q is 104, the insertion loss at a critical coupling amounts to 0.035dB and at Q=103 it is 0.33dB. For filter with a band of 100 MHz external Q-factor is 400. At the same unloaded Q-factor, BPF insertion loss will equal 0.17dB and 1.46dB, respectively. Hence it is evident that the transition to HTS materials in creating BPF is expedient for narrow-band filters.

Estimate of the resonator own Q-factor of the fabricated filter gives value of the order 5000. The study of the microwave field in the resonator showed that there is a redistribution of current density, namely, it is reduced on the walls of the waveguide, and it increases on the edges of insert "windows". It is this effect determines the appropriateness of the use of HTS E-plane inserts.

We also note one technical contradiction, manifested in the manufacture of BPF with HTS insert. To ensure good electrical and thermal contact, we need to tightly clamp the insert between the half housing waveguide parts. This leads to mechanical stresses in the brittle MgO inserts under cooling and frequent cracking of inserts. It is necessary to find new solutions to the problem marked.

References

1.R.R. Mansour, A. Zybura, “Superconducting Millimeter-Wave E-Plane Filters”, IEEE Trans. Microwave Theory Tech., Vol.39, No.9, pp.1588-1492, 1991.

2.L. Han, Y. Chen, Y. Wang, Q. Cheng, S. Yang and P. Wu, Design and Performance of Waveguide E-Plane HTSC Insert Filters, 1992 IEEE IMS-Digest, IF2 1-4, pp. 912-916.

3.R. Vahldieck, J. Bornemann, F. Arndt, D. Grauerholz, “Optimized Waveguide E-Plane Metal Insert Filters for Millimeter-Wave Applications”, IEEE Trans. Microwave Theory Tech., Vol.31, No.1, pp.65-69, 1983.

4.F. Arndt, J. Bornemann, D. Grauerholz, R. Vahldieck, Theory and Design of Low-Insertion Loss Fin-Line Filters, IEEE Trans. Microwave Theory Tech., VOL. MTT-30, NO. 2, FEBRUARY 1982 pp. 155-163.

5.Y.-C. Shih, T. Itoh, and L. Q. Bui, Computer-Aided Design of Millimeter-Wave E-Plane Filters, IEEE Trans. Microwave Theory Tech., VOL. M~-31, NO. 2, FEBRUARY 1983, pp. 135-142.

6.J.L. Matthae, L. Young, E.M.T. Jones, Microwave Filters, Impedance-Matching Networks, and Coupling Structures. McGraw-Hill Co., 1968.

7.V.N. Skresanov, V.V. Glamazdin, N.T. Cherpak “The Novel Approach to Coupled Mode Parameters Recovery from Microwave Resonator Amplitude-Frequency Response”, European Microwave Conference Proceedings (EuMC 2011), pp. 826-829, 2011.

663