A new WLR method based on model parameter analysis

Abstract

An approach which identifies key disturbances which cause bipolar model parametric variations and correlation is presented. Model parameters, expressed as a function of technological data, are analyzed through a straightforward correlation study. Combining these results the key process variables fluctuation which causes circuit performance changes can be isolated. This methodology which links model parameter and circuit performance to process quantities, is a benefit to both device manufacturers and end users.