Application of Monte Carlo methods to the analysis of circuit sensitivity due to component tolerance

Abstract

A Turbo-Pascal program has been developed that will aid a circuit designer to determine which of a set of "real" circuits will best match the associated "ideal" circuit or in choosing components for a given circuit. Study one examined the Sallen and Key second-order active low-pass circuit. Wn (mean), Wn (standard deviation), Q (mean), and Q (standard deviation) were examined as a function of Q desired, distribution type and component tolerance. The sensitivity of the circuit was calculated both numerically and through program execution. In Study two a set of reference graphs were generated to compare several second order low pass, high pass, and band pass filter classes as a function of component distribution and tolerance. Study three compares the behavior of several orders of low pass MFM (Maximally Flat Magnitude) Leapfrog filters. Amax (maximum gain), Wmax (frequency of maximum gain), and the 3dB frequency were examined. (Abstract shortened with permission of author.)