An Improved Design Method For Stepped Line Microwave Filters with Broad Stop Bands

The design problem for filters attenuating signals with frequencies outside the useful (pass) bands can be in many cases reduced to the problem of designing low pass or band-stop filters with sufficiently large bandwidths. It follows from the fact that some elements of complex electronic systems, as for example the radiating elements of antenna arrays, have the characteristics similar to those shown in Figure 1. Among various possible implementations of band-stop filters, those implemented as a cascade connection of transmission line sections—the so-called stepped transmission line band-stop filters [1, 2]—are especially suitable for many applications. These filters are rather simple to manufacture, and their small crosssections allow you to put them in place of specified sections of the transmission line. Among all analytical design methods, the most general is the one described in [1 4]. Its essential feature is the use of R-transformer as a prototype circuit together with the classical Darlington-Riblet method [1 5]. Unfortunately, the filters designed according to this last method may be difficult to manufacture in some cases because of the large spread of the characteristic impedances among the line sections. This follows from the fact that corresponding insertion loss function is formed by an appropriate choice of characteristic impedances of sections having equal electrical lengths. Consequently, for a fixed number of sections, the range of characteristic impedances increases when the requirements imposed on the relative stop band of the filter become more restrictive. Therefore, the aim of this paper is to present the new approach that makes possible the design of band-stop filters with required insertion loss function (including filters with broad stop bands), which can be easily implemented in the prescribed technology. It can be obtained by limiting the range of characteristic impedances by application of optimization methods with constraints. Due to the fact that both characteristic impedances and electrical lengths of the sections are variables during the optimization process, this method belongs to the class of amplitude-phase methods. The optimization process starts with an initial approximation found by means of an appropriate analytical method [1 4]. Naturally, the efficiency of the optimization strategy strongly depends on the quality of initial approximation. For example, analytical The technique described here emphasizes proper setup of initial parameters before applying optimization in the well-known microwave EDA tools