A numerical method for the synthesis of distributed RC networks
McKnight, Randy Sherwood
Burrus, C. Sidney
Master of Science
Synthesis of RC distributed networks is considered in this thesis. The behavior of these networks is formulated as a differential equation in terms of impedance. Using this behavioral equation and a boundary condition, an optimization problem is constructed so as to yield a solution that approximates a specified impedance characteristic. The technique is demonstrated for several example network configurations. By assuming an initial structure, a gradient algorithm is employed to solve the optimization problem. Two S-plane representations of the impedance, one along the oj-axis and the other along the a-axis, were used in the solution. Both representations afforded rapid convergence, with the o-axis examples achieving more accurate structures in less time. These examples clearly demonstrate the effectiveness and practicability of this method.