Modeling And Parameter Extraction For High Electron Mobility Transistors

With the raising requirements for products,the radio frequency(RF)and microwave circuit design has become increasingly complicated,so that the circuit designers have to utilize the computer aided design(CAD)softwares to design and optimize their circuits.To shorten the design cycle and improve the first-pass design success,The CAD softwores are required to provede fast and accurate models for transistors,which is the basis of accurte simulation of the circuit performance.In RF and microwave field,high electron mobility transistors(HEMTs)have been popular used in the design of the large-signal nonlinear blocks,such as power amplifiers,mixes,etc.However,the accuracy of nonlinear models for HEMT is not satisfactory,especially for the prediction of the high-order products of current-voltage(I-V)characteristics,resulting in the difference between the measurement and simulation.To solve this problem,the I-V nonlinear modeling for GaAs pHEMT has been studied.The major research is as follows:Firstly,the small signal equivalent circuit model of GaAs pHEMTs has been developed.The fourteen parameters have been extracted step by step using the measured DC and S parameters.The values of parasitic and intrinsic parameters have been determined.To validite the effectiveness,the measured and simulated S paremeters in different bias conditions have been compared.The modeled small signal circuit provides good basis for large signal modeling.Secondly,a nonlinear model for HEMTs considering the output current Ids and its high-order derivatives has been proposed.The proposed model divides the entire operating region into several regions and chooses different existing empirical models in each region to acquire high accuracy.In order to keep the model continuity,the artificial neural network(ANN)techniques are employed to smoothly link the boundaries.The validity of the proposed model has been demonstrated by comparing the measured and modeled Ids and its high-order derivatives of a pHEMT.Moreover,this method can also be employed to model high-order derivatives of HEMTs with other technology.Finally,an artificial neural network(ANN)based model has been proposed to describe the characteristics of output current and its high-order derivatives of HEMTs over a wide temperature range.The temperature is considered as one of the input vectors of the ANN.To verify the validity of the proposed model,the nonlinear I-V model for pHEMT with accurate high-order derivatives(at least up to the third order)has been developed within the temperature range of-25?~125?.What more,the results show that the proposed ANN model presents a better agreement of high-order derivatives than the popularly used Angelov model.