Statistical circuit analysis based on SABER

Abstract

It is extremely useful for electrical engineers to predict design performance as component parameters deviate from the criterion value randomly. This work can be handled by a circuit simulation software, SABER, which covers powerful statistical analysis including Monte Carlo analysis and worst-case analysis. In the simulation environment of SABER, the circuit statistical analysis at randomly selected device parameters under certain tolerance bounds is illustrated in detail with a flyback convertor and an active band pass filter as examples. Moreover, the simulation wave and statistical summary are presented. Based on the statistical analysis results, the robustness of the circuit is appraised and the proportion of finished product in the batch production is forecasted. Furthermore, the acceptable tolerance bounds on the device parameters are obtained. The work in this paper is a practical guide to high robustness circuit design.