Computer-aided design of distributed RC networks

In this paper, synthesis of distributed RC networks using nonlinear programming is investigated. The Fletcher-Powell minimization procedure is used to minimize the integral squared error. At the same time, the decreasing of the error in the Chebyshev sense is examined in each iterative cycle. One of the main considerations in applying nonlinear programming to a design problem is to find an initial design such that, by iterating it in the frequency range of interest, the iteration is guaranteed to converge to a minimum within a reasonably short computing time. Two methods for finding such an initial design are presented. One is to make an approximation of the given network function in a transformed complex frequency domain. Both approximations are obtained by means of orthogonal functions. The methods presented in this paper may be used for practical distributed RC network applications. Examples are given to illustrate the methods.