A rigorous exposition of the LEMMA method for analog and mixed-signal testing.

Abstract

The linear error-mechanism modeling technique is an effective tool for testing analog and mixed-signal devices which minimizes the number of measurements required to characterize the static transfer function of a circuit by determining a small number of parameters of a linear error model and then predicting the entire response error. This work focuses on optimizing the linear error-mechanism model algorithm (LEMMA), introducing novel refinements which are shown to improve its performance significantly. We outline the implementation of the algorithm in a tutorial manner, paying due consideration to the underlying theory where required.