Statistical analysis method of gan device process parameters based on large signal equivalent circuit model

Abstract

A process parameter analysis method based on the GaN device equivalent circuit model is proposed. The analysis method includes: step 1: establish a GaN device small-signal equivalent circuit model, and extract small-signal model parameters; step 2: establish a GaN device large-signal, etc. The effective circuit model is used to extract the large signal model parameters, namely the nonlinear current source model parameters and the nonlinear capacitance model parameters; Step 3: Targeting the measured microwave characteristics of the device, tuning and optimizing the large signal model parameters; Step 4: Based on the established large signal model parameters The signal model extracts process parameters of multiple batches of GaN devices, and performs statistical analysis on the process parameters. The statistical analysis method of the process parameters of the above-mentioned GaN device model first establishes the GaN device small-signal equivalent circuit model, and then establishes the GaN device large-signal equivalent circuit model associated with the process parameters, and finally obtains the statistical distribution of process parameters through multi-batch device modeling , useful for device yield analysis and process parameter optimization.

Description

technical field [0001] The invention relates to the technical field of GaN HEMT (gallium nitride high electron mobility transistor) devices, in particular to a statistical analysis method for GaN device process parameters based on a large-signal equivalent circuit model. Background technique [0002] Gallium Nitride High Electron Mobility Transistor (GaNHEMT) has extremely important applications in microwave and millimeter wave solid-state power circuits due to its high frequency and high power density. The current mainstream method of circuit design is usually based on the device model that describes the characteristics of the device under small-signal working conditions and large-signal working conditions in the form of an equivalent circuit, so the device model is the premise of using the device for circuit design. [0003] However, due to unintentional doping and process parameter fluctuations in the device preparation process, it will affect the consistency of device pe...

AI Technical Summary

Problems solved by technology

[0003] However, due to unintentional doping and process parameter fluctuations in the device preparation process, it will affect the consistency of device performance, thereby affecting the yield of circuit design, so it is necessary to guide the analysis of circuit yield by establishing a statistical model

Traditional statistical methods are based on the small signal model or some large signal model parameters, so the accuracy is insufficient

Moreover, it is impossible to obtain specific statistical analysis of process parameters through large-signal statistical models to guide device yield design and process parameter optimization

Images