диафрагмированные волноводные фильтры / f0025289-b883-48e4-aa52-8ba839bd311d

Abstract — Substrate integrated waveguide (SIW) technique provides a low-profile, low-cost and low-weight scheme while maintaining high performance, which is particularly useful for airborne and satellite system applications. In this paper, two K- band 4th-degree and 5th-degree oversized SIW cavity filters with symmetrical inductive post-wall irises are designed and realized on the basis of a conventional microwave substrate of Rogers RT/Duroid 6002 by arrays of via. A finite transmission zero is assigned by the interaction between the fundamental TE101 mode and the higher-order TE301 mode to improve the attenuation in the transmit band. A good agreement is observed between the simulation results and measurement result even though there is some difference due to the fabrication error.

Index Terms — Substrate Integrated Waveguide (SIW), Filter, Transmission Zero (TZ), Satellite and Airborne Communications, Millimeter-Wave Technology

performance is poor due to the very strong couplings and spurious resonances. Figure 1 shows the field theory-based simulation results for a 4th-order directly-coupled symmetrical

I. INTRODUCTION

Satellite communication systems and airborne equipment require low-cost, low-profile and compact size microwave and millimeter-wave filters with low insertion loss [1-2]. Conventional waveguide filter has very small insertion loss; however, it is bulky and very difficult to integrate with active components. On the other hand, microstrip filter has a small size while its insertion loss is high. Substrate Integrated Waveguide (SIW) technology has been demonstrated to provide an attractive solution to low-cost microwave and millimeter-wave waveguide components and systems. In particular, this technology allows the use of standard PCB or LTCC process for designing and making low-cost waveguide components which can easily be integrated with planar circuits on the same substrate [3-5].

The Rx filter of a Ka-band satellite ground terminal has an intermediate fractional bandwidth of about 9% and a wide stopband performance, especially in the Tx band where sufficient stopband attenuation should be provided. Directlycoupled SIW filter is a good candidate because it is very suitable for wide passband applications and easy to design and fabricate. As for the low insertion loss in the passband, four SIW cavities are enough to meet the requirement on the selectivity which is not so stringent. However, the stopband

Fig.1 Simulation Results for a 4th-degree directly-coupled SIW filter

SIW filter with dual posts as coupling elements. The minimum insertion loss in the passband is about 0.54dB and the filter performs well around the passband of (19.2-21.2) GHz. However, its performance deteriorates when they operate at the higher frequency. The stopband attenuation is reduced significantly towards higher frequency; especially at the Tx band of (29.5-30) GHz the attenuation is small due to the spurious pass band. The reason for this phenomenon is that the dual posts can be viewed as the coupling inverters only in a narrow frequency band and an increasing amount of power is carried across the coupling sections, even before the onset of the second-fundamental-resonance passband which also contribute to the poor stopband attenuation.

Many techniques have been developed to improve the stopband performance of conventional rectangular waveguide filters [6]. They can be classified into two categories: (1) Improve the performance of coupling elements. For example, good stopband performance of E-plane filter can be obtained by reducing the distance between the metal insert and

waveguide sidewalls with a thick metal insert or several metal inserts rather than a single one so that the propagation of modes along the coupling sections is suppressed up to a higher frequency resulting in an improved stopband attenuation; (2) Widen the distance between the fundamental resonance and the spurious. The resonators are dimensioned to resonate at the same fundamental resonance and not all simultaneously resonate at any higher frequency. In most cases, capacitive elements formed by the E-plane discontinuities are used to load the resonator, for example, ridged-waveguide resonators, E-plane stepped impedance resonators (SIRs), etc.

These techniques are not suitable for SIW filters. As we know, transmission zeros (TZs) can be used to improve the selectivity and stopband attenuation. In general the implementation of transmission zeros in the insertion loss response of a microwave filter can be obtained using the well known “extracted pole” technique [7] or by introducing couplings between nonadjacent resonators (cross couplings) [8]. Using the extracted pole technique, one cavity can be designed to produce at the same time one transmission pole (return loss minima) and one transmission zero; however, the TZ can not be far away from the desired passband. Using cross couplings, what is actually done from a physical point of view in order to generate a transmission zero is to provide two paths to the signal with the proper phasing and magnitude so that cancellation can occur at a given frequency. One path is the main signal path while the other is the cross coupling, and the stronger the cross coupling the closer the zeros to the filter bandpass.

The actual implementation of the cross or bypass coupling is either physical or modal. In the case of physical cross coupling, a physical element such as a coupling aperture in waveguide filters or fringing fields in planar filters is used. This technique has the advantage that the coupling element is a physically identifiable and adjustable element. However, the cross-coupled filters with physical coupling are not practical for wideband applications because of the physical dimension limitations. Also, the finite TZs (FTZs) may not be far away from the passband because the physical coupling can not be very small. An alternative approach is the use of other modes, propagating or evanescent [9-13], as separate paths for energy flow, which is very suitable for wideband waveguide filters with direct-coupled-resonators. In this paper, a TZ generated by the nonphysical cross coupling of higher order modes is used to improve the stopband attenuation of a K-band SIW filter.

II. PRINCIPLE AND DESIGN

In an oversized SIW cavity directly excited by input/output microstrip lines, a TZ can be generated due to the excitation of higher order modes in the cavity when the coupling coefficients to the spurious resonance are markedly larger

than those to the main one. The fundamentaland secondorder modes provide two separate paths for the signal between the two ports of an oversized cavity. Naturally, other modes which may be excited at the ports will provide additional paths, but they are assumed to be weak enough to be neglected. The position of TZ can be decided by the following relationship

where ωz’ is the generalized angular frequency of the TZ. J1’ and J2’ are the generalized coupling admittances between the source/load and the main resonance. J3’and J4’ are the generalized coupling admittances between the source/load and the spurious resonance. Bsp’ is the generalized constant susceptances of spurious resonance [13].

TE101 and TE301 modes are used for the symmetrical cavity in the width dimension. Figure 2 shows the magnetic

Fig.2 Magnetic field distribution of TE101 and TE301 modes

field distribution of the TE101/TE301 mode. Obviously, all the couplings between the input/output microstrip lines and TE101/TE301 modes have the same type (positive) and the coupling between the input/output microstrip lines and the TE101 mode is significantly stronger than that between the input/output microstrip lines and the TE301 mode. Thus if the

TE101 mode is viewed as the spurious resonance and the TE301 mode as the main resonance, a TZ below the resonance of

TE301 mode can be generated and this TZ is very close to the resonance of TE301 mode because the coupling between the

input/output microstrip lines and the TE101 mode is much stronger than that between the input/output microstrip lines

and the TE301. Figure 3 shows the frequency response of the oversized SIW TE101/TE301 cavity directly excited by

input/output microstrip lines. A TZ between the two resonances is generated and this TZ is very close to the resonance of TE301 mode.

Fig.3 Response for the oversized SIW TE101/TE301 mode cavity

The oversized TE101/TE301 mode cavity can be used to design a microwave filter with a defined passband and

transmission zero location. At first the dimensions of the SIW cavity is determined by setting the resonant frequency of TE101 mode equal to the center frequency of the desired passband. Then we proceed to adjust the parameters of cavity for the position of the transmission zeros. Because the ratio of the coupling between the input/output microstrip lines and the TE101 mode to that between the input/output microstrip lines

and the TE301 mode is not easy to change in the symmetrical oversized SIW cavity in the width direction, the position zero

can be tuned only by changing the resonant frequency of the spurious resonance (TE101 mode). Once the cavity is nearly optimized, we can fine tune the cavity again in order to obtain the proper zero location and we conclude with the final adjustment of the cavity parameters. Next the sizes of the coupling post-wall irises can be determined by the classical process [14]. At last the whole filter can be tuned to meet the requirement. This process is very fast essentially due to the fact that the transmission zeros generated are not close enough to the edge of the bandpass to significantly interact with the transmission poles of the filters (insertion loss minima) so that a pure Chebyshev filter response can still be achieved.

III. FABRICATION AND MEASUREMENT

A 4th-degree and a 5th-degree oversized SIW cavity filters with the passband of (19.2-21.2) GHz and TZ at (29.5-30) GHz are both designed and realized on Rogers RT/ Duroid 6002 with the height of 20mil by following the procedure above. The material has a dielectric constant of 2.94 and a loss tangent of 0.0012. Dimensions of the two filters are shown in Table I.

An HP 8510C vector network analyzer and Anritsu Wiltron 3680K test fixture are used to measure the two filters. A thru-reflect-line (TRL) calibration is performed in order to remove the effect of the test fixture. Figure 4 shows their

TABLE I

DIMENSIONS OF TWO OVERSIZED SIW CAVITY FILTERS

simulation and measurement results, along with their layouts. The minimum insertion loss in the passband for the full-wave simulation of the 4th- and 5th-degree filters is about 0.6dB and 0.7dB, respectively. The minimum insertion loss in the passband for the measurement of the 4th- and 5th degree filters is about 0.8dB and 1.1dB, respectively. Simulated stopband attenuation in the Tx frequency band of 29.5-30 GHz for the 4th- and 5th-degree filters is better than 45dB and 60dB, respectively. Measured stopband attenuation achieves approximately 32dB and 35dB for the 4th- and 5th-degree filters, respectively. There is also a slight upward frequency shift in the measured response, particularly for the 5th-degree filter. In fact, the fabrication error leads to the difference between the simulation and measurement results.

(a) Simulated and measured narrowband responses of 4th-degree filter

(b) Simulated and measured wideband responses of 4th-degree filter

(c) Simulated and measured narrowband responses of 5th-degree filter

(d) Simulated and measured wideband responses of 5th-degree filter

Fig.4 Simulation and Measurement Results for two SIW filters

VI. CONCLUSION

In this paper, two K-band oversized SIW cavity filters are designed and realized on a conventional dielectric substrate (Rogers RT/Duroid 6002) for a passband from 19.2-21.2 GHz. A transmission-zero (TZ) generated by the cross coupling of higher-order modes is used to improve the stopband attenuation, especially at the Tx band of 29.5-30 GHz. The simulated and measured results show that the SIW filters may be the choice for the Rx filter in a satellite communication ground terminal if the insertion loss in the passband and the stopband attenuation at the Tx band are further improved. In the future work, our effort will be made on the generation of multiple TZs used to improve the stopband attenuation.

ACKNOWLEDGEMENT

The authors express their gratitude to J. Gauthier, R. Brassard, and S. Dube, all with Poly-Grames Research Center, École Polytechnique de Montréal, Quebec, Canada, for their technical assistance in fabricating the experimental prototypes.

Special thanks also to Mario Caron and Richard Paiement who gave many helpful suggestions and encouragements as well as early critical review of this work. Finally, the authors would like to acknowledge that this work was in part funded by, and performed under contract for the Government of Canada, represented by the Communications Research Centre (CRC) Canada. This work also received a financial support from Natural Sciences and Engineering Research Council (NSERC).

REFERENCES

[1]M. Lecouve, P. Jany, E. Kerherve, N. Boutheiller, and F. Marc, “Genetic Algorithm Optimization for Evanescent Mode Waveguide Filter Design,” in IEEE Int. Symp. Circuits and Systems, Geneva., Switzerland, May 28-31, 2000.

[2]S. Christopher; G. R. Rewankar, V.A. Abid Hussain, and N. Rama Prasad, “Rugged receiver for airborne application”, in IEEE Int. Symp. Circuits Microwaves, Radar and Wireless Communications, May 20-22, 2002, vol.2, Pp: 663 – 668.

[3]K. Wu, D. Deslandes, and Y. Cassivi, “The substrate integrated circuits

– a new concept for high-frequency electronics and optoeletronics”,

presented at Telecommunications in Modern Satellite, Cable and Broadcasting Service, 2003. TELSIKS 2003. 6th International Conference on, vol. 1, Oct. 1-3, 2003 Pages:P-III - P-X.

[4]D. Deslandes and K. Wu, “Single-substrate integration techniques for planar circuits and waveguide filters”, IEEE Trans. Microwave Theory Tech., vol. 51, no.2, pp593-596, Feb. 2003.

[5]D. Deslandes, K. Wu, “Integrated Microstrip and Rectangular Waveguide in Planar Form”, IEEE Microwave wireless Component Letters, vol. 11, no.2, pp.68-70, Feb. 2001.

[6]D. Budimir, “Optimized E-plane bandpass filters with improved stopband performance”, IEEE Trans. Microwave Theory Tech., vol.45, no.2, pp212-219, 1997.

[7]J. D. Rhodes and R. J. Cameron, “General extracted pole synthesis

technique with applications to low-loss TE011 mode filters,” IEEE Trans. Microwave Theory Tech., vol. MTT-28, no. 9, pp.1018-1028, 1980.

[8]R. Levy, “Filters with single transmission zeros at real and imaginary frequencies,” IEEE Trans. Microwave Theory Tech., vol. MTT-24, pp.172-181, 1976.

[9]F. Arndt, T. Duschak, U. Papziner, and P. Roalppe, “Asymmetric iris coupled cavity filters with stopband poles,” in IEEE MTT-S Int. Microwave Symp. Dig., 1990, pp. 215-218.

[10]K. Iguchi, M. Tsuji, and H. Shigesawa, “Negative coupling between TE and TE modes for use in evanescent-mode bandpass filters and their field-theoretic CAD,” in IEEE MTT-S Int. Microwave Symp. Dig., 1994, pp. 727-730.

[11]U. Rosenberg, W. Hagele, “Consideration of parasitic bypass coupling in overmoded cavity filter designs,” IEEE Trans. Microw. Theory Tech., vol. 42, no. 7, pp. 1301-1306, 1994.

[12]M. Guglielmi, F. Montauti, L. Pellegrini, and P. Arcioni, “Implementing transmission zeros in the inductive-window bandpass filters,” IEEE Trans. Microw. Theory Tech., vol. 43, no. 8, pp. 19111915, 1995.

[13]S. Amari, U. Rosenberg, “Characteristics of Cross (Bypass) Coupling Through Higher/Lower Order Modes and Their Applications in Elliptic Filter Design,” IEEE Trans. Microwave Theory Tech., vol. 53, no. 10, pp3135-3141, 2005.

[14]J.-S. Hong, M.J .Lancaster, Microstrip filter for RF/Microwave applications. New York: John Wiley & Sons, 2001.