Frequency-scalable nonlinear behavioral transistor model from single frequency X-parameters based on time-reversal transformation properties (INVITED)

This paper presents a powerful new method that generates a frequency-scalable nonlinear simulation model from single-frequency large-signal transistor X-parameter data. The method is based on a novel orthogonal identification (direct extraction) of current source and charge source contributions to the spectrally rich port currents under large-signal conditions. Explicit decomposition formulae, applied entirely in the frequency domain, are derived in terms of sensitivity functions at pairs of large-signal operating points related to one-another by time-reversal transformation. The method is applied and validated with respect to data from a measurement-based model of a pHEMT transistor. It is demonstrated that the scalable model can predict the nonlinear performance of the transistor over several orders of magnitude in frequency, all from X-parameters at a single fundamental frequency.