Abstract

This paper presents an efficient, physics-based approach to evaluate the large-signal (LS) parametric sensitivity of microwave devices operated in (quasi) periodic conditions. The approach is based on the combined use of a two-dimensional drift-diffusion physical model and of an intermediate circuit-oriented large-signal model. The LS parametric sensitivity is then exploited to optimize the doping profile of a GaAs MESFET so as to minimize the third-order intermodulation products (IMP) in a simple large-signal amplifier. An example of yield-oriented statistical analysis of device performances with respect to random variations of technological parameters is also presented.