Workshop Chair: Dr. W.Grabinski
Technical Program Coordinator: Prof. H.Iwai

Abstract: 
MOSFET models for circuit simulation have recorded important advances in physical and numerical accuracy over the last two generations. In particular the latest generation of models, PSP, BSIM5, EKV3, HiSIM, ACM, etc., implement for the first time physically continuous equations in all regions of operation based on the surface potential. The problem is that these models have become increasingly complex and a wide gap exists between the MOSFET representation used for designing circuits and the one used for verifying the circuits. This paper describes a modeling methodology that can bridge the gap between design and verification. Sound analog design starts with an exploration phase rather than applying “design by SPICE.” Design exploration is an essential step for tradeoffs among the different requirements of an analog circuit such as gain, bandwidth, power and area, and, leads to a suitable circuit topology with sized transistors. A simple yet sufficiently accurate model is derived based on the charge sheet approximation, similar to the EKV and ACM models, that describes operation from subthreshold to strong inversion and can be stripped down to a form only slightly more complex than the Shichman-Hodges (Level=1) quadratic model. Circuit design based on this simplified model can be carried out using the gm/ID methodology if a handful of key parameters of the model are adjusted to fit measured I-V and C-V characteristics. The model, its parameters and their extraction, and, its application to the design of opamps are presented. The key merit of the described model is that the initial circuit design will closely match SPICE for a given process, independent of the simulation model used.