An Intrinsic Small-Signal Equivalent Circuit Model for AlGaN/GaN HEMT Considering the Momentum Balance Equation

Y Pei,Fujiang Lin,Eugeny A Ryndin,Amgad A Al-Saman,Chengong Yin

doi:10.1109/jeds.2021.3124327

Abstract

High-frequency dispersion of the AlGaN/GaN HEMT output resistance has been investigated. Using the electron momentum balance equation, we demonstrated that the deviation of the dynamic drift velocity from its static value could be considered as the main reason for the high-frequency dispersion of the transistor output resistance. Moreover, a new intrinsic small-signal equivalent circuit has been proposed to consider the output resistance dispersion. Comparison with measured s-parameters have validated the proposed approach.

Highlights

TTE small-signal modelling of microwave GaN HEMT devices is a research topic of high importance in microwave circuit design and device characterization [1, 2]
The small-signal model is required to reproduce the GaN HEMT transistor behavior, and it is considered as the first step to developing the large-signal and noise models [3-5]
Two approaches of small-signal modelling coexist: 1) the electrical approach, which is based on the equivalent circuit models, and 2) the behavioral approach, which is based on the artificial neural networks (ANNs) [3]

Summary

INTRODUCTION

TTE small-signal modelling of microwave GaN HEMT devices is a research topic of high importance in microwave circuit design and device characterization [1, 2]. Even though the behavioral approach gives an accurate reproducing of the S-parameters measurement, it has no connection with device physics and circuitry. This makes such models less preferable for RF and microwave designers. At MIMC design, especially at high frequency or mm-wave, the matching itself becomes an issue, and any inaccuracy of the modeled S22 may lead to poor output performance of the design. To overcome such difficulties and physically explain such high-frequency dispersion, the equivalent circuit's intrinsic part will be investigated

MODEL DESCRIPTION

MODEL VALIDATION

Findings

CONCLUSIONS