An Improved Noise Modeling Method Using a Quasi-Physical Zone Division Model for AlGaN/GaN HEMTs

Accurate characterization of transistor noise performance is significant for low-noise amplifier (LNA) design. The conventional empirical noise model contains too many fitting parameters and thus relies on a large number of on-wafer noise measurements, especially the ambient temperature effects modeling. In this article, an improved noise modeling method using the quasi-physical zone division (QPZD) theory is proposed. The boundary potential calculation in the traditional modeling method is improved by considering the effects of the charge depletion region when the transistor is in the ON-state, which can result in at least two orders of magnitude improvement in the accuracy of the noise power spectral density (PSD). With this improvement, the noise performance can be accurately predicted without adding any fitting parameters in the noise PSD models. Furthermore, ambient temperature modeling of the noise performance is realized by considering the ambient temperature effects of drain–source current
${I}_{\text {ds}}$
with a modified
${V}_{\text {off}}$
model and some other significant intrinsic parameters. Verification shows that the proposed QPZD noise model with the improved boundary potential calculation method can well predict the noise performance and scattering parameters at different ambient temperatures. The method of this study can be applied to physical-based noise modeling of GaN high-electron-mobility transistors (HEMTs) and LNA design.