Microwave Filters for Communication Systems: Fundamentals, Design and Applications

CHAPTER 9

RECONFIGURATION OF THE FOLDED COUPLING MATRIX

In Chapters 7 and 8 we described two different techniques for synthesizing the canonical folded-array coupling matrix from the transfer and the reflection polynomials of the desired filter function. The folded array has several advantages for the filter designer:

• Layout possibilities are relatively simple.
• Positive and negative couplings can be implemented to realize advanced filter characteristics.
• The maximum (N – 2) number of finite-position transmission zeros can be realized if required.
• The diagonal couplings, necessary for asymmetric characteristics, can be implemented in some cases.

However, if the filter is to be realized in dual-mode technology (with two orthogonal resonant modes supported in the same physical resonator), a disadvantage becomes apparent. For a folded-array realization, the input and output of the filter are in the same physical cavity. This imposes a limit on the input/output isolation that can be achieved with bandpass filters, typically 25 dB for the dual-TE_11n (Transverse Electric) mode in cylindrical resonator cavities or dual-TE_10n mode in square resonator cavities. Consequently, for dual-mode filters, the input and output couplings must occur in separate resonators. This chapter is devoted to the methods for the similarity transformations of the folded coupling array to realize a wide range of topologies suitable for dual-mode filter networks.

Figure 9.1 Sixth-degree network (a) cross-coupled ...