Ambiguity groups and testability

An efficient method has been developed for determining component ambiguity groups which arise in analog circuit testing. The method makes use of the sensitivity model of the circuit. The ambiguity groupings are shown to depend on the test points selected and the measurement accuracy and are therefore a useful tool for determining where to add or delete test points. The concept of ambiguity groups can be used to refine the testability measure of a circuit. An example that demonstrates the effects of ambiguity groups on parameter estimation and performance prediction is presented. >