An Improved Nonlinear Capacitance Model for GaN HEMTs Based on the Angelov Model

To accurately model the nonlinear characteristics of the intrinsic capacitance of Gallium Nitride High Electron Mobility Transistors (GaN HEMTs), this paper proposes an improved nonlinear capacitance model based on the traditional Angelov model. The experiments employed GaN HEMTs fabricated by United Monolithic Semiconductors (UMS) using the GH15-10 process. These devices feature a uniform gate length of 150 nm, with gate widths designed as 2 × 10 μm, 4 × 20 μm, 6 × 30 μm, and 8 × 40 μm, respectively (gate width is calculated as "number of gate fingers × unit gate width"). In this study, two types of devices with gate widths of 2 × 10 μm and 4 × 20 μm were selected to extract intrinsic capacitance parameters, which were then substituted into the traditional Angelov model and the proposed improved model, respectively. By comparing the coefficient of determination (R²) and Root Mean Square Error (RMSE) of the two models, it is found that the improved model has a significantly higher degree of agreement with the experimental measurement dates and is more suitable for the quantitative characterization of the nonlinear capacitance characteristics of GaN HEMTs. This research provides support for the accurate application of device models in the design of high-frequency circuits in subsequent studies.