FET STATISTICAL MODELING USING PARAMETER ORTHOGONALIZATION

A new method for representing the statistical variation of FET equival ent circuit parameters (ECP's) is presented. This method utilizes a st atistical technique known as principal components and provides an effi cient method for statistically representing the means, standard deviat ions, and correlations of the FET ECP's. The technique can easily be i mplemented into commercial CAD simulators resulting in FET variation s imulations that are more accurate than existing methods. Appropriate s tatistical tests for determination of equivalence between simulated an d measured FET parameter distributions is also discussed. Both the mod eling methodology and statistical testing were demonstrated using both scattering and noise parameters for 300 mu m low-noise GaAs FET's.