Optimum test voltages for breakdown experiments

Abstract

When performing breakdown measurements on an electrical system, the test voltages have to be chosen carefully in order to yield maximum information about the breakdown characteristic of the system. The better the test voltages are chosen the smaller is the experimental effort for getting a predefined result quality. In this investigation the question "What are the best test voltages in a breakdown experiment?" is addressed. A general method is presented that allows to determine the optimum test voltage U/sub N+1/, when N breakdown experiments at voltages U/sub 1/,...U/sub N/ have already been performed. With this method a sequential determination of the optimum test voltages is possible. The usage of optimum test voltages lead to a minimization of the number of breakdown experiments when a predefined result quality has to be reached, or the result quality is maximized for a given number of breakdown experiments.