In this paper, it is shown empirically and through simulation that transistor mismatch due to random dopant fluctuation is a function of the well and halo design of the transistor, and that, contrary to conventional expectation, low-threshold transistors can have larger mismatch than higher threshold transistors. The complex dependence of mismatch on well and halo profiles suggests the need for the extension of the conventional Pelgrom approach to characterizing mismatch for a given technology and also suggests means of optimizing mismatch for analog applications. A set of screening criteria for mismatch data analysis are presented to verify that conclusions drawn from the standard deviation of a distribution may be properly applied.
No takes yet. Share an insight, caveat, or question.
Hook et al. (2010) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: