GaN HEMT Model Research Articles

Behavioral-level modeling of GaN HEMT small-signal intrinsic noise based on DE-SVR algorithm

The Differential Evolution-Support Vector Regression (DE-SVR) algorithm is designed to model the small-signal intrinsic noise behavior of GaN HEMT. It not only overcomes the local minimization shortcoming of the Back Propagation (BP) algorithm, but also uses the DE (Differential Evolution) algorithm to obtain the best parameter c (punishment factor) and parameter g (variance of kernel function) of the Support Vector Regression (SVR) algorithm. In order to validate the superiority of the DE-SVR algorithm, the experiment compares the modeling effects of BP algorithm, SVR algorithm, and DE-SVR algorithm in modeling the small-signal intrinsic noise model of GaN HEMT. The experimental results show that there are obvious advantages for the DE-SVR algorithm in modeling the small-signal intrinsic noise characteristics of GaN HEMT.

Temperature dependent analytical model for transfer characteristics of GaN HEMTs with AlxInyGazN barrier layer

We present a temperature dependent analytical model for GaN HEMT with quaternary alloy AlxInyGazN barrier layer. HEMTs are also known as Heterojunction field effect transistors (HFETs). Change in the ambient temperature of device and the self-heating effects vary its threshold voltage and performance. We compare our calculated data with the available experimental data and find that the computed outcomes agree with the experimental observations existing in literature. Thereby, validating our claim that the proposed model might find use in the simulation of quaternary alloy based AlInGaN devices for various applications.

High‐frequency noise modeling of GAN HEMT with double recessed barrier layer

AbstractA physics‐based high‐frequency noise model of AlGaN/GaN HEMT with double recessed barrier layer (DRBL) is presented. A conventional GaN HEMT model is established on TCAD and verified by measurement results. A noise model of DRBL GaN HEMT up to 40 GHz is investigated based on the established model. The results show that DRBL HEMT reduces the high frequency noise and the reduction becomes more obvious as the frequency increases. Further investigation about noise performance is studied by changing the depth and length of the recessed regions. The results show that DRBL HEMT obtains the best noise performance when the length and depth are 400 nm and 9 nm respectively and has application potential in the development of millimeter waves low noise devices.

A comparative study on the accuracy of small-signal equivalent circuit modeling for large gate periphery GaN HEMT with different source to drain length and gate width

In this paper, extrinsic and intrinsic parameters were extracted from the experimental S-parameters using 16- and 22-element small signal equivalent circuit model for large gate periphery GaN HEMT with different source to drain length and gate width over 500 MHz to 26 GHz range. The extraction using 22-element model shows good consistency to measured S-parameters and the maximum gain, stability factor, delta factor and K-factor with less percentage error than 16-element model. As the source to drain length is increasing, the accuracy of parameter extraction is enhanced for 22-element model. Specially for stability factor and K-factor the percentage error was reduced drastically (10.42%–4.74% and 8.13%–0.83% respectively). For increased number of fingers, the 22-element model shows more accuracy. Also, TCAD (Technology Computer-Aided Design) device simulations were done and transfer characteristic were calibrated with measured data. The simulated s-parameters were then calibrated with the experimental extrinsic parameters.

An Accurate Neural Network-Based Consistent Gate Charge Model for GaN HEMTs by Refining Intrinsic Capacitances

Neural network-based capacitance models are accurate, but some of them are not charge-conservative. In this work, a novel consistent gate charge model for GaN high electron mobility transistors is presented based on neural networks. The equivalent circuit parameters are extracted using the multiobjective gray wolf optimizer-based hybrid method, which improves the accuracy of parameter extraction. To obtain more reliable data sets for accurate neural network-based modeling, the outliers in the extracted intrinsic capacitances are automatically detected and removed using the isolation forest technique. The gate charge is obtained by integrating the capacitances with the voltages at different temperatures. A neural network is used to model the bias- and temperature-dependent gate charges, and the intrinsic capacitance formulation is obtained by taking the partial derivative of the gate charge function with respect to the voltages. The proposed model is charge-conservative and requires no transcapacitances. The large-signal model is implemented in the Advanced Design System and verified by small- and large-signal measurements. Good agreement is obtained between the measurements and simulations.

Modeling and Analysis of Double Channel GaN HEMTs Using a Physics-Based Analytical Model

In this work, we report the development of a new physics-based analytical model for current and charge characteristics of Double Channel (DC) Gallium Nitride High Electron Mobility Transistors (GaN-HEMTs). The model has at its core the self-consistent calculation of the charge densities, for both upper and lower channels, obtained from a solution of the Schrodinger and Poisson equations. Fermi-Dirac (FD) distribution together with 2D density of states is used for mobile carrier statistics in both the channels. Furthermore, drift-diffusion transport is used to compute the channel current using charge densities at channel extremities. Finally, the model is validated against TCAD simulation and experimental data for a DC-GaN-HEMT. The model, by virtue of its fully physics-based nature, precisely captures the charge screening effect in the lower channel without invoking any empirical clamping functions, unlike prior models, and also possesses the feature of being geometrically scalable.

Analysis and Suppression of High Speed Dv/Dt Induced False Turn-on in GaN HEMT Phase-Leg Topology

Gallium nitride high electron mobility transistor (GaN HEMT) is liable to gate false turn-on problem when the gate crosstalk voltage exceeds its threshold voltage in the widely adopted phase-leg topology due to its low threshold voltage and high switching speed. Without considering the gate loop stray inductance, gate internal resistance, nonlinearity of parasitic capacitances and power loop stray parameters, traditional false turn-on analytical method is insufficient to support accurate analysis. And it has been found that GaN HEMT gate-source parasitic capacitance C <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">gs</sub> previously assumed constant is otherwise highly nonlinear and has strong impacts on the gate crosstalk voltage. This paper has measured C <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">gs</sub> by vector network analyzer and constructed an accurate nonlinear model of C <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">gs</sub> , based on which an accurate GaN HEMT behavior model is further fulfilled. The accuracy of the proposed behavior model has been verified by large amounts of experiment results. The proposed GaN HEMT model is used to accurately calculate gate crosstalk voltage and switching losses. Besides, false turn-on induced extra loss has been calculated and is adopted as a criterion to evaluate the severity of false turn-on and optimal design method for false turn-on suppression has been detailed further.

An Accurate Datasheet-Based Full-Characteristics Analytical Model of GaN HEMTs for Deadtime Optimization

The gallium nitride high electron mobility transistors (GaN HEMTs) are a superior candidate for the new-generation power electronics systems with higher efficiency and power density. However, due to the unique reverse characteristics, the reverse voltage drop of GaN HEMTs is much higher than that of diode. The deadtime loss in GaN-based bridge converters will be comparable to switching losses if the deadtime is not optimized. To optimize the deadtime for higher efficiency, this article proposes an accurate analytical model of GaN HEMTs, including circuit's parasitic inductances, the nonlinear capacitances, the unique reverse characteristics, etc. Taking a GaN-based synchronous buck converter as the example, the proposed model is realized, which fully uses the datasheet to avoid additional experiments. In order to accurately measure the switching current for validation, a novel parasitics-based current measurement method is proposed. The proposed model is verified by simulation in LTspice and experiment, and good agreement is shown. Based on the accurate analytical model, the deadtime is optimized for different load currents to improve the efficiency within the full load range. Compared with the fixed deadtime of 15 ns, the increase of efficiency can be up to 8%. This work will promote the high-frequency application of GaN HEMTs.

140 W급 전력 증폭 소자의 병렬 확장형 GaN HEMT 대신호 모델

In this study, a scalable GaN HEMT large-signal model for 140-W power devices is presented. The large-signal model of an 80-finger GaN HEMT includes eight unit device models in parallel, and the large-signal model of the unit device is based on the AMCAD–FET model. The unit device includes a 10-finger GaN HEMT, a gate pad, and a drain pad. Parameters for the unit GaN HEMT model were extracted using an IVCAD program of AMCAD and a Keysight PNA-X network analyzer. The unit GaN HEMT model was verified through on-wafer load-pull measurements, and the large-signal GaN HEMT model of the large-sized 80-finger device was verified through load-pull measurements of the packaged device.

An RF Approach to Modelling Gallium Nitride Power Devices Using Parasitic Extraction

This paper begins with a comprehensive review into the existing GaN device models. Secondly, it identifies the need for a more accurate GaN switching model. A simple practical process based on radio frequency techniques using Vector Network Analyser is introduced in this paper as an original contribution. It was applied to extract the impedances of the GaN device to develop an efficient behavioural model. The switching behaviour of the model was validated using both simulation and real time double pulse test experiments at 500 V, 15 A conditions. The proposed model is much easier for power designers to handle, without the need for knowledge about the physics or geometry of the device. The proposed model for Transphorm GaN HEMT was found to be 95.2% more accurate when compared to the existing LT-Spice manufacturer model. This work additionally highlights the need to adopt established RF techniques into power electronics to reduce the learning curve while dealing with these novel high-speed switching devices.

GaN-Based Megahertz Single-Phase Inverter With a Hybrid TCM Control Method for High Efficiency and High-Power Density

Compared with traditional frequency limitation method DCM and QCFTCM for TCM-based inverter, this article proposes a hybrid TCM method, which can achieve both full range ZVS and frequency variation limitation. An accurate resonant transition model is established and the switching frequency variation range can be obtained for TCM. The hybrid control can be implemented according to the variation ratio γ. In order to determine the optimal γ, an offline optimization algorithm based on loss estimation is proposed. Therefore, an accurate analytical loss model for GaN HEMTs considering the nonlinear junction capacitance is given for TCM and hybrid TCM. The loss model shows that the total losses of hybrid TCM inverter are reduced compared with TCM. Moreover, an interleaved GaN-based megahertz single-phase inverter with digital control is demonstrated with 120 W/in <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> power density. The aforementioned benefits are experimentally verified in the optimal hybrid TCM. With this hybrid control, the measured peak efficiency is 98%. Compared with pure TCM mode, there is a 70% reduction in the switching frequency and the related efficiency is 0.8% higher at light loads. Higher power density (135 W/in <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> ) can be achieved in hybrid TCM through further increasing the switching frequency compared with pure TCM.

Large-Signal Modeling of GaN HEMTs Using Hybrid GA-ANN, PSO-SVR, and GPR-Based Approaches

This article presents an extensive study and demonstration of efficient electrothermal large-signal GaN HEMT modeling approaches based on combined techniques of Genetic Algorithm (GA) with Artificial Neural Networks (ANN), and Particle Swarm optimization (PSO) with Support Vector Regression (SVR). Another promising Gaussian Process Regression (GPR) based large-signal modeling approach is also explored and presented. The GA-ANN addresses the typical problem of local minima associated with the backpropagation (BP) based ANN. The GA successfully aids in the determination of optimal initial values for BP-ANN and enables it to find a unique optimal solution after subsequent of iterations with higher rate of convergence. This is also achieved using PSO-SVR with lower optimization variables. The developed modeling techniques are demonstrated and used to simulate the gate and drain currents of a 2-mm GaN device. All the models are relatively simple, practical, and easy to implement. The gate and drain currents models are embedded in an equivalent large-signal circuit's model and built in Advanced Design System (ADS) software. The implemented model is validated by large-signal measurements and very good fitting results have been obtained. The model also showed an accurate simulation for a nonlinear power amplifier with very good computational speed and convergence.

A Combined Broadband Model for GaN HEMTs in Admittance Domain Based on Canonical Piecewise Linear Functions

In this article, a new large-signal transistor model is proposed. The model combines the advantage of both compact modeling and behavioral modeling techniques, resulting in a new modeling methodology. The intrinsic part of the model is based on the black-box poly-harmonic distortion (PHD) framework, which is implemented using the canonical piecewise linear (CPL) functions. Thus, the new intrinsic model can cover a broad operating frequency band, across a wide range of input powers, using a compact set of model parameters. While at each fixed operating frequency and input power, the model can also accurately predict device behavior over the complete Smith chart. For the extrinsic model, the traditional extrinsic parasitic network is used, enabling access to the intrinsic device plane, as required for waveform engineering. This work examines intrinsic admittance domain models implemented using both the linear and quadratic poly-harmonic expansions. An <inline-formula> <tex-math notation="LaTeX">$8 \times 125\,\,\mu \text{m}$ </tex-math></inline-formula> GaN HEMT device with a 0.25-<inline-formula> <tex-math notation="LaTeX">$\mu \text{m}$ </tex-math></inline-formula> gate feature size and a commercial 10-W GaN HEMT is used during the simulation and experimental model validation. Model implementations show excellent interpolation and extrapolation capability over the frequency band across a range of input power levels in both the frequency domain and the time domain.

Scaled GaN-HEMT Large-Signal Model Based on EM Simulation

This paper presents a scaled GaN-HEMT large-signal model based on EM simulation. A large-signal model of the 10-finger GaN-HEMT consists of a large-signal model of the two-finger GaN-HEMT and an equivalent circuit of the interconnection circuit. The equivalent circuit of the interconnection circuit was extracted according to the EM simulation results. The large-signal model for the two-finger device is based on the conventional Angelov channel current model. The large-signal model for the 10-finger device was verified through load-pull measurement. The 10-finger GaN-HEMT produced an output power of about 20 W for both simulation and load-pull measurements.

Consistent Surface-Potential-Based Modeling of Drain and Gate Currents in AlGaN/GaN HEMTs

In this article, the gate current in AlGaN/GaN high-electron mobility transistors is modeled in a surface potential-based compact model. The thermionic emission, the Poole–Frenkel emission, and the Fowler–Nordheim tunneling are the dominant mechanisms for the gate current in the forward- and reverse-bias regions. These conduction mechanisms are modeled within the framework of the ASM-GaN compact model, which is a physics-based industry-standard model for GaN HEMTs, hence yielding a consistent model for the drain and gate currents. The proposed model captures the gate voltage, drain voltage, temperature, and gate-length dependencies of the gate current. The results of dc gate-leakage measurements of two GaN HEMT, differing only in terms of gate length, over a wide range of temperature, showing these current-conduction mechanisms, are presented, and the proposed model is validated accordingly. The developed gate current model, implemented in Verilog-A, is in excellent agreement with the experimental data.

Modeling the Impact of the High-Field Region on the $C-V$ Characteristics in GaN HEMTs

In this article, we illustrate the impact of the high electric field region and the effects of this has on the capacitance–voltage characteristics of a GaN HEMT device. Such effects arise due to a significant spike in the electric field near the drain-side edge of the gate electrode. The presence of the high electric field has a direct impact on the intrinsic capacitances of the device. We present a physics-based compact model for the gate charge and intrinsic capacitances, which captures these effects. TCAD simulations are performed to analyze the underlying principles of such phenomena while providing validation for the proposed model. The advanced SPICE model for GaN HEMTs (ASM-GaN-HEMT) is adopted and modified to include the effects of the high electric field region. The simulation results are in excellent agreement to the TCAD and the measured data. The model is compared to simulations that neglect such effects in order to illustrate the importance of capturing the impact of high-field region on the intrinsic device capacitances.

A streamlined drain-lag model for GaN HEMTs based on pulsed S-parameter measurements

AbstractAccurately and efficiently modeling the drain-lag effects is crucial in nonlinear large-signal modeling for Gallium Nitride high electron mobility transistors. In this paper, a simplified yet accurate drain-lag model based on an industry standard large-signal model, i.e., the Chalmers (Angelov) model, extracted by means of pulsed S-parameter measurements, is presented. Instead of a complex nonlinear drain-lag description, only four constant parameters of the proposed drain-lag model need to be determined to accurately describe the large impacts of the drain-lag effects, e.g., drain-source current slump, typical kink observed in pulsed IV curves, and degradation of the output power. The extraction procedure of the parameters is based on pulsed S-parameter measurements, which allow to freeze traps and isolate the trapping effects from self-heating. It is also shown that the model can very accurately predict the load pull performance over a wide range of drain bias voltages. Finally, the large-signal network analyzer measurements at low frequency are used to further verify the proposed drain-lag model in the prediction of the output current in time domain under large-signal condition.

Differential evolution based small signal modeling of GaN HEMT

In this article differential evolution based method of small signal modeling of GAN HEMT has been investigated. The method uses a unique search space exploration strategy to obtain optimized values of intrinsic and extrinsic elements pertaining to compact small signal model from extracted equivalent circuit elements and measured S-parameter data. Effectiveness of the method has been illustrated by comparing the measured S-parameter data of a 4 × 0.1 × 75 μm2 GaN/SiC HEMT in the frequency range of 1 to 30 GHz wherein modeled and measured data are in good agreement.

ASM GaN: Industry Standard Model for GaN RF and Power Devices—Part 1: DC, CV, and RF Model

We present the latest developments in Advance SPICE Model for GaN (ASM GaN) HEMTs in this paper. The ASM GaN model has been recently selected as an industry-standard compact model for GaN radio frequency (RF) and power devices. The core surface-potential calculation and the modeling of real device effects in this model are presented. We discuss the details of the nonlinear access region model and enhancement in this model to include a physical dependence on barrier thickness. We also present the novel model feature of configurable field-plate modeling and discuss the extraction procedure for the same. New results with the ASM GaN model on high-frequency and enhancement-mode GaN power devices are also presented.

ASM GaN: Industry Standard Model for GaN RF and Power Devices—Part-II: Modeling of Charge Trapping

Because of charge trapping in GaN HEMTs, dc characteristics of these devices are not representative of high-frequency operation. The advanced spice model GaN model presented in Part I of this paper is combined with a Shockley–Reed–Hall-based trap model, yielding a comprehensive FET model for GaN HEMTs which can accurately model GaN devices exhibiting trapping-related dispersion effects. Measurement results of the dc and pulsed output and transfer characteristics of a commercially available GaN HEMT are presented, trapping in the device is modeled, and excellent fit to the measured data is shown. This paper presents an accurate model of trapping which is validated for eight different quiescent bias points of pulse measurements, with quiescent drain voltage ranging from 5 to 20 V and quiescent gate voltage ranging from −2.8 to −3.8 V, and a large range of gate and drain voltages to which the device was pulsed in the pulse measurements and at which the device was measured in the dc measurements, with gate voltage ranging from −4 to 0.4 V and drain voltage ranging from 0 to 40 V. This paper also presents high-frequency (10 GHz) large-signal RF validation of the model for optimal complex load condition.