14 March 2023 in IEEE Transactions on Electron Devices
The thermal analysis of GaN-based high electron-mobility transistors (HEMTs) is carried out from steady-state and transient aspects, respectively. An improved closed-form nonlinear thermal model based on device structure is proposed. In the steady-state case, the plane layer is added on the basis of the original partition, and the effect of thermal boundary resistance on channel temperature is also considered. In addition, the operating temperature of each gate finger is predicted based on the thermal coupling, and the nonlinear thermal model is extended to the transient by combining the thermal time constant. The validity of our presented model is verified by comparing the simulation results with the experimental data. Furthermore, this model is also embedded in the surface-potential-based current model to verify the self-heating effect in the circuits.
22 February 2023 in IEEE Transactions on Electron Devices
In our experimental preparation of perovskite solar cells (PSCs), when the thickness of the electron transport layer (ETL) was changed from thin to thick, the current–voltage ( ${I}$ – ${V}$ ) characteristics showed the changing laws of S-shaped kinks appearance, S-shaped kinks disappearance, and S-shaped kinks appearance. A lumped-parameter equivalent circuit model is proposed to reveal the intrinsic physical significances of the S-shaped kinks for this variation laws. The explicit solution of the model is derived using the deformed difference-microvariation (DM) principle combined with the region method, and the model parameters are extracted efficiently through the firefly optimization algorithm (FA). The proposed analytical solution enables the model to be compactly implemented in photovoltaic devices and circuit simulators, and the model is expected to point the direction for process optimization of devices.
17 August 2022 in IEEE Transactions on Electron Devices
An analytical dc model of AlGaN/GaN and AlGaAs/GaAs-based high-electron mobility transistors (HEMTs) is proposed, while the charge accumulated in the barrier layer at high gate biases voltage is taken into account. To break through the bottleneck problem of solving model accurately and efficiently, an improved difference-microvariation (DM) explicit algorithm is developed to solve the complex transcendental equations of Fermi level and surface potential in the physical model, including two important subbands in a triangular potential well. Compared with the existing analytical surface potential-based algorithms, the DM algorithm quantifies the high-order component and obtains a more precise solution. The universality and accuracy of the model and DM solutions are verified against numerical results and experimental data under a wide range of operation region, which proves that the model combined with DM method is expected to become the suitable simulation tools for generic heterojunction HEMTs.