A Hybrid Computer-Aided Tuning Method for Microwave Filters

Abstract

Abstract-A hybrid tuning method for microwave filters is presented in this paper. This novel tuning technique is based on the combination of the Cauchy method and aggressive space mapping (ASM) technique. Cauchy method is applied to determine the characteristic polynomials of the filter's response, then the parameters (coupling matrix) of the low-pass prototype is extracted from the characteristic polynomials. The aggressive space mapping is used to optimized the fine model to guarantee that each step of a tuning is always in the right direction. The validity is verified by two examples. One deals with the fourresonator cross-coupled filter and the other one is an direct coupled six-resonator filter.

Diagnosis of Coupled Resonator Bandpass Filters Using Vf and Optimization Method

“…1(b), the simulated response of filter 1 is compared to the extraction one calculated by the extracted CM in Eq. (14). Very good agreement between the simulated and extracted responses can be observed.…”

“…Early filter diagnosis techniques can be traced back to the 1970s and 1980s [1,2]. In the past years, interest is growing on the filter diagnosis methods for extracting the coupling matrix of microwave filters from the measurements (or simulations) with losses [3][4][5][6][7][8][9][10][11][12][13][14][15]. These filter diagnosis or coupling matrix (CM) extraction techniques require two sorts: 1) CM extraction techniques are based on optimization (as in [3][4][5]) for matching the measured scattering parameters, whose elements are the problem unknowns.…”

“…The optimization techniques are either greatly influenced by the initial values assigned to the variables, number of the resonators and topology of the filters, or are time consuming; and 2) CM extraction techniques are based on a polynomial model, which matches the measured (or the simulated) parameters and then extract the CM of the filter with a specified topology (by using well-known established methods as in [16]). Polynomial models may refer to the Y -parameters as in [6][7][8][9] or to the scattering parameters (S-parameters) as in [10][11][12][13][14][15]. When using the S-parameters models to diagnose a filter including losses, the loss effect associated to each resonator must be removed in advance using either the optimization method (as in [12,13]) or a modified frequency transformation including unloaded Q in [11].…”

Self‐learning tuning of coaxial cavity filter by using the poles and residues of admittance function

“…Lastly, the poles and residues of the Y-parameters are obtained by using the partial fraction expansion method. The phase shift of the dual port caused by the high-order mode and transmission line is as follows [17]:…”

“…Lastly, the poles and residues of the Y ‐parameters are obtained by using the partial fraction expansion method. The phase shift of the dual port caused by the high‐order mode and transmission line is as follows [17]: where φ 01 and φ 02 denote the phase loading caused by higher‐order modes; denotes the phase shift introduced by the transmission line at a sampling frequency ω , α 1 and α 2 denote the equivalent electrical length of input and output transmission lines, respectively; and ω 0 denotes the centre frequency. The relationship between the measured and the ideal S ‐parameters is as follows: …”