On large-signal modeling of GaN HEMTs: past, development and future

In the past few decades,circuits based on gallium nitride high elec-tron mobility transistor(GaN HEMT)have demonstrated exceptional potential in a wide range of high-power and high-frequency applica-tions,such as the new generation mobile communications,object de-tection and consumer electronics,etc.As a critical intermediary be-tween GaN HEMT devices and circuit-level applications,GaN HEMT large-signal models play a pivotal role in the design,application and development of GaN HEMT devices and circuits.This review pro-vides an in-depth examination of the advancements in GaN HEMT large-signal modeling in recent decades.Detailed and comprehensive coverage of various aspects of GaN HEMT large-signal model was of-fered,including large-signal measurement setups,classical formula-tion methods,model classification and non-ideal effects,etc.In order to better serve follow-up researches,this review also explored poten-tial future directions for the development of GaN HEMT large-signal modeling.

MOS Model 20 Based RF-SOI LDMOS Large-Signal Modeling

A novel large-signal equivalent circuit model of RF-SOI LDMOS based on Philips MOS Model 20 （MM20） is presented. The weak avalanche effect and the power dissipation caused by self-heating are described. The RF parasitic elements are extracted directly from measured S-parameters with analytical methods. Their final values can be obtained quickly and accurately through the necessary optimization. The model is validated in DC,AC small-signal,and large-signal analyses for an RF-SOI LDMOS of 20-fingers （channel mask length, L = 1μm,finger width, W = 50μm） gate with high resistivity substrate and body-contact. Excellent agreement is achieved between simulated and measured results for DC, S- parameters （10MHz-0.01GHz）, and power characteristics, which shows our model is accurate and reliable. MM20 is improved for RF-SOI LDMOS large-signal applications. This model has been implemented in Verilog-A using the ADS circuit simulator （hpeesofsim）.

Effect of junction temperature on the large-signal properties of a 94 GHz silicon based double-drift region impact avalanche transit time device

The authors have developed a large-signal simulation technique extending an in-house small-signal simulation code for analyzing a 94 GHz double-drift region impact avalanche transit time device based on silicon with a non-sinusoidal voltage excitation and studied the effect of junction temperature between 300 and 550 K on the large-signal characteristics of the device for both continuous wave （CW） and pulsed modes of operation. Results show that the large-signal RF power output of the device in both CW and pulsed modes increases with the increase of voltage modulation factor up to 60%, but decreases sharply with further increase of voltage modulation factor for a particular junction temperature; while the same parameter increases with the increase of junction temperature for a particular voltage modulation factor. Heat sinks made of copper and type-IIA diamond are designed to carry out the steady-state and transient thermal analysis of the device operating in CW and pulsed modes respectively. Authors have adopted Olson＇s method to carry out the transient analysis of the device, which clearly establishes the superiority of type-IIA diamond over copper as the heat sink material of the device from the standpoint of the undesirable effect of frequency chirping due to thermal transients in the pulsed mode.

Si/Si C-based DD hetero-structure IMPATTs as MM-wave power-source:a generalized large-signal analysis

A full-scale, self-consistent, non-linear, large-signal model of double-drift hetero-structure IMPATT diode with general doping profile is derived. This newly developed model, for the first time, has been used to analyze the large-signal characteristics of hexagonal SiC-based double-drift IMPATT diode. Considering the fabrication feasibility, the authors have studied the large-signal characteristics of Si/SiC-based hetero-structure devices. Under small-voltage modulation （~ 2%, i.e. small-signal conditions） results are in good agreement with calculations done using a linearised small-signal model. The large-signal values of the diode＇s negative conductance （5 × 106 S/m2）, susceptance （10.4 × 107 S/m2）, average breakdown voltage （207.6 V）, and power generating efficiency （15%, RF power： 25.0 W at 94 GHz） are obtained as a function of oscillation amplitude （50% of DC breakdown voltage） for a fixed average current density. The large-signal calculations exhibit power and efficiency saturation for large-signal （〉 50%） voltage modulation and thereafter decrease gradually with further increasing voltage-modulation. This generalized large-signal formulation is applicable for all types of IMPATT structures with distributed and narrow avalanche zones. The simulator is made more realistic by incorporating the space-charge effects, realistic field and temperature dependent material parameters in Si and SiC. The electric field snap-shots and the large-signal impedance and admittance of the diode with current excitation are expressed in closed loop form. This study will act as a guide for researchers to fabricate a high-power Si/SiC-based IMPATT for possible application in high-power MM-wave communication systems.

Analysis and prediction of loudspeaker large-signal symptoms

Nonlinear lumped-parameter force factor Bl(x), stiffness Kms(x) and inductance Le(x) of electrodynamic loudspeakers change frequency responses and generate some nonlinear effects for large stimulus: harmonic and intermodulation distortion, DC component in diaphragm displacement, instability of vibration and jumping effects. By modeling the nonlinear system under large-signal conditions, relationship between the nonlinear parameters and large-signal behavior can be revealed and help to provide guidance to diagnose loudspeakers. Agreement between the measured and predicted responses of a real loudspeaker validates the modeling and enables new methods for loudspeaker diagnosis.

Large-signal characterizations of DDR IMPATT devices based on group Ⅲ–Ⅴ semiconductors at millimeter-wave and terahertz frequencies

Large-signal （L-S） characterizations of double-drift region （DDR） impact avalanche transit time （IM- PATT） devices based on group III-V semiconductors such as wurtzite （Wz） GaN, GaAs and InP have been carried out at both millimeter-wave （mm-wave） and terahertz （THz） frequency bands. A L-S simulation technique based on a non-sinusoidal voltage excitation （NSVE） model developed by the authors has been used to obtain the high frequency properties of the above mentioned devices. The effect of band-to-band tunneling on the L-S properties of the device at different mm-wave and THz frequencies are also investigated. Similar studies are also carried out for DDR IMPATTs based on the most popular semiconductor material, i.e. Si, for the sake of comparison. A compara- tive study of the devices based on conventional semiconductor materials （i.e. GaAs, InP and Si） with those based on Wz-GaN shows significantly better performance capabilities of the latter at both mm-wave and THz frequencies.

Large-signal characterization of DDR silicon IMPATTs operating in millimeter-wave and terahertz regime

The authors have carried out the large-signal characterization ofsilicon-based double-drift region （DDR） impact avalanche transit time （IMPATT） devices designed to operate up to 0.5 THz using a large-signal simulation method developed by the authors based on non-sinusoidal voltage excitation. The effect of band-to-band tunneling as well as parasitic series resistance on the large-signal properties of DDR Si IMPATTs have also been studied at different mm-wave and THz frequencies. Large-signal simulation results show that DDR Si IMPATT is capable of delivering peak RF power of 633.69 mW with 7.95% conversion efficiency at 94 GHz for 50% voltage modulation, whereas peak RF power output and efficiency fall to 81.08 mW and 2.01% respectively at 0.5 THz for same voltage modulation. The simulation results are compared with the experimental results and are found to be in close agreement.

A symbolically defined InP double heterojunction bipolar transistor large-signal model

A self-built accurate and flexible large-signal model based on an analysis of the characteristics of InP double heterojunction bipolar transistors （DHBTs） is implemented as a seven-port symbolically defined device （SDD） in Agilent ADS. The model accounts for most physical phenomena incluuing the self-heating effect, Kirk effect, soft knee effect, base collector capacitance and collector transit time. The validity and the accuracy of the large-signal model are assessed by comparing the simulation with the measurement of DC, multi-bias small signal S parameters for InP DHBTs.

Large-signal SPICE model for depletion-type silicon ring modulators

We present an accurate, easy-to-use large-signal SPICE circuit model for depletion-type silicon ring modulators(Si RMs). Our model includes both the electrical and optical characteristics of the Si RM and consists of circuit elements whose values change depending on modulation voltages. The accuracy of our model is confirmed by comparing the SPICE simulation results of 25 Gb/s non-return-to-zero(NRZ) modulation with the measurement. The model is used for performance optimization of monolithically integrated Si photonic NRZ and pulse-amplitudemodulation 4 transmitters in the standard SPICE circuit design environment.

Large-signal modeling of three-phase dual active bridge converters for electromagnetic transient analysis in DC grids

The three-phase dual active bridge(3 p-DAB)converter is widely considered in next-generation DC grid applications.As for traditional AC grids,the successful integration of power electronic converters in DC grids requires accurate time-domain system-level studies.As demonstrated in the existing literature,the development and efficient implementation of large-signal models of 3 pDAB converters are not trivial.In this paper,a generalized average model is developed,which enables system-level simulation of DC grids with 3 p-DAB converters in electromagnetic transient type(EMT-type)programs.The proposed model is rigorously compared with alternative modeling techniques:ideal-model,switching-function and state-space averaging.It is concluded that the generalized average model provides an optimal solution when accuracy of transient response,reduction in computation time,and wideband response factors are considered.

MEXTRAM model based SiGe HBT large-signal modeling

An improved large-signal equivalent-circuit model for SiGe HBTs based on the MEXTRAM model （level 504.5） is proposed. The proposed model takes into account the soft knee effect. The model keeps the main features of the MEXTRAM model even though some simplifications have been made in the equivalent circuit topology. This model is validated in DC and AC analyses for SiGe HBTs fabricated with 0.35-μm BiCMOS technology, 1 × 8 μm^2 emitter area. Good agreement is achieved between the measured and modeled results for DC and S-parameters （from 50 MHz to 20 GHz）, which shows that the proposed model is accurate and reliable. The model has been implemented in Verilog-A using the ADS circuit simulator.

A millimeter wave large-signal model of GaAs planar Schottky varactor diodes

A millimeter wave large-signal model of GaAs planar Schottky varactor diodes based on a physical analysis is presented.The model consists of nonlinear resistances and capacitances of the junction region and external parasitic parameters.By analyzing the characteristics of the diode under reverse and forward bias,an extraction procedure of all of the parameters is addressed.To validate the newly proposed model,the PSVDs were fabricated based on a planar process and were measured using an automatic network analyzer.Measurement shows that the model exactly represents the behavior of GaAs PSVDs under a wide bias condition from -10 to 0.6 V and for frequencies up to 40 GHz.

RF CMOS modeling:a novel empirical large-signal model for an RF-MOSFET

A novel empirical model for large-signal modeling of an RF-MOSFET is proposed. The proposed model is validated in the DC, AC, small-signal and large-signal characteristics of a 32-finger nMOSFET fabricated in SMIC＇s 0.18 μm RF CMOS technology. The power dissipation caused by self-heating is described. Excellent agreement is achieved between simulation and measurement for DC, S-parameters （50 MHz-40 GHz）, and power characteristics, which shows that our model is accurate and reliable.

A complete and accurate surface-potential based large-signal model for compound semiconductor HEMTs

A complete and accurate surface potential based large-signal model for compound semiconductor HEMTs is presented. A surface potential equation resembling the one used in conventional MOSFET models is achieved. The analytic solutions from the traditional surface potential theory that developed in MOSFET models are inherited. For core model derivation, a novel method is used to realize a direct application of the standard surfacepotentialmodelofMOSFETsforHEMTmodeling,withoutbreakingthemathematicstructure.Thehigh-order derivatives of I–V /C–V remain continuous, making the model suitable for RF large-signal applications. Furthermore, the self-heating effects and the transconductance dispersion are also modelled. The model has been verified through comparison with measured DC IV, Gummel symmetry test, CV, minimum noise figure, small-signal Sparameters up to 66 GHz and single-tone input power sweep at 29 GHz for a 475 m0.1 m InGaAs/GaAs power pHEMT, fabricated at a commercial foundry.

CLOSE-LOOP DESIGN OF PULSE POWER SUPPLY FOR ELECTRICAL DISCHARGE MACHINING

The dynamic performance and the stability are essential for a system. A new circuit topology used for electrical discharge machining (EDM) power and made up of complex-pulse (voltage-pulse and current-pulse) is presented. The large-signal model of it is also derived. Based on the comprehensive analysis of the system model, the compensator is designed to make the system match better and to improve its dynamic performance and the stability under perturbations. Finally, the design methods and the analysis are verified by simulation and experimental results.

Theoretical and numerical studies on a planar gyrotron with transverse energy extraction

In this paper, we study the planar gyrotron theoretically and numerically. Applying the large-signal theory to the planar gyrotron, the wave equation of electric field and the equation of motion of an electron are simultaneously solved to obtain some characteristic parameters, such as the phase-space plot of electrons, working frequency, startup time, electronic efficiency, and output power of the device. To verify the formulations used in this paper, three-dimensional particle simula- tions are performed on the same device, and the numerical results accord well with those obtained by using the large-signal theory. Theoretical and numerical results show that the electronic efficiency can reach 21% for the prototype of the planar gyrotron working at the frequency of 0.81 THz.

CAD OF MILLIMETER WAVE OSCILLATORS WITH THE FDHB TECHNIQUE

In this paper the modelling, analysis and optimization of millimeter wave oscillatorsare investigated by using the a frequency-domain harmonic balance technique (FDHB), where theexternal-circuit impedances looking outside from the active device are calculated with a combinedtechnique of modes expansion, Galerkin, and collocation methods. The optimization results arein agreement with the experimental ones, which show the reliability of the presented model andoptimization.

Theoretical analysis and an improvement method of the bias effect on the linearity of RF linear power amplifiers

Based on the Gummel-Poon model of BJT, the change of the DC bias as a function of the AC input signal in RF linear power amplifiers is theoretically derived, so that the linearity of different DC bias circuits can be interpreted and compared. According to the analysis results, a quantitative adaptive DC bias circuit is proposed, which can improve the linearity and efficiency. From the simulation and test results, we draw conclusions on how to improve the design of linear power amplifier.

An improved nonlinear model of HEMTs with independent transconductance tail-off fitting

We present an improved large-signal device model of GaAs/GaN HEMTs, amenable for use in com- mercial nonlinear simulators. The proposed model includes a new exponential function to independently control the transconductance compression/tail-off behaviors. The main advantage of this model is to provide a simple and coherent description of the bias-dependent drain current （Ⅰ-V） that is valid in all regions of operation. All aspects of the model are validated for 0.25μm gate-length GaAs and GaN HEMT processes. The simulation results of DC/pulsed Ⅰ-V, RF large-signal power and intermodulation distortion products show excellent agreement with the measured data.

An improved large signal model of InP HEMTs

An improved large signal model for InP HEMTs is proposed in this paper.The channel current and charge model equations are constructed based on the Angelov model equations.Both the equations for channel current and gate charge models were all continuous and high order drivable,and the proposed gate charge model satisfied the charge conservation.For the strong leakage induced barrier reduction effect of InP HEMTs,the Angelov current model equations are improved.The channel current model could fit DC performance of devices.A 2×25μm×70 nm InP HEMT device is used to demonstrate the extraction and validation of the model,in which the model has predicted the DC I–V,C–V and bias related S parameters accurately.