This paper describes a new, general and accurate, large signal GaAs FET model for nonlinear (e.g. harmonic balance) circuit simulation, its fast and unambiguous construction (model generation) by explicit calculations applied to the raw device data, and the adaptive, automated data acquisition system used to characterize the device. The model implementation in a harmonic balance simulator, the model generation procedure, and the automated data acquisition system form an efficient, practical, commercially available package. for state-of-the-art nonlinear circuit design.

This paper derives the theory and method of the two dimensional discrete TL and CTL models including high frequency skin effect for an arbitrary shaped plane pair in a multiple dielectric layer structure. The null or peak of the S parameter frequency response represents the test port interaction with the resonant standing wave of these planes at that frequency. The resultant S parameter data of these models can be condensed into a simpler N port equivalent circuit to represent a larger hierarchical power and ground plane network for fast simulation.

The new measurement-based HP FET Model [l] greatly improves large-signal accuracy because the model nonlinearities are explicitly constructed from device data. To achieve this accuracy, a software product based on the I-IP IC-CAP program was developed to acquire FET data and package it for use in a nonlinear circuit simulator. This paper describes a n adaptive, automated data acquisition system (ADAS) for device on-wafer measurements and its application to the HP FET Model generation process. The ADAS solves the problem of automatically determining the entire, device-specific, safe operating region of the FET from only data sheet information. The system adaptively chooses the bias points at which to measure, collects DC and small-signal S-parameter data, and sends it to the HP FET RilodelGenerator-where it is packaged for use with the new HP FET Model.Because the techniques involved are very general, they can be easily applied for characterizing other classes of devices and to provide FET data for more conventional parameter extraction techniques.'Hewlett Packard Co., h4WTD. Sant,a Rosa, CA 95403

No abstract