CN105844042A - Large signal statistical model modeling method for gallium nitride high electron mobility transistor - Google Patents

Large signal statistical model modeling method for gallium nitride high electron mobility transistor Download PDF

Info

Publication number
CN105844042A
CN105844042A CN201610203716.0A CN201610203716A CN105844042A CN 105844042 A CN105844042 A CN 105844042A CN 201610203716 A CN201610203716 A CN 201610203716A CN 105844042 A CN105844042 A CN 105844042A
Authority
CN
China
Prior art keywords
electron mobility
parameter
high electron
large signal
modeling method
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Withdrawn
Application number
CN201610203716.0A
Other languages
Chinese (zh)
Inventor
徐跃杭
陈志凯
徐锐敏
延波
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
University of Electronic Science and Technology of China
Original Assignee
University of Electronic Science and Technology of China
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by University of Electronic Science and Technology of China filed Critical University of Electronic Science and Technology of China
Priority to CN201610203716.0A priority Critical patent/CN105844042A/en
Publication of CN105844042A publication Critical patent/CN105844042A/en
Priority to CN201710120505.5A priority patent/CN106845025B/en
Withdrawn legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/30—Circuit design
    • G06F30/36—Circuit design at the analogue level
    • G06F30/367—Design verification, e.g. using simulation, simulation program with integrated circuit emphasis [SPICE], direct methods or relaxation methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/30—Circuit design
    • G06F30/39—Circuit design at the physical level
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/30—Circuit design
    • G06F30/39—Circuit design at the physical level
    • G06F30/398—Design verification or optimisation, e.g. using design rule check [DRC], layout versus schematics [LVS] or finite element methods [FEM]

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Geometry (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Microelectronics & Electronic Packaging (AREA)
  • Junction Field-Effect Transistors (AREA)

Abstract

The invention provides a large signal statistical model modeling method for a gallium nitride high electron mobility transistor. The method comprises the following steps: testing a plurality of gallium nitride high electron mobility transistors in each batch, and acquiring current-voltage characteristics of the gallium nitride high electron mobility transistors; allowing the current-voltage characteristics to be used in large signal equivalent circuit models of the gallium nitride high electron mobility transistors, and extracting large signal equivalent circuit parameters about the gallium nitride high electron mobility transistors; and using a response curve method to build a large signal statistical model according to the large signal equivalent circuit parameters. The modeling method provided by the invention can reduce the amount of data, and avoid the problem of non-convergence when an abnormal value occurs, so that the statistical model obtained through the modeling method can accurately reflect process change conditions of different devices, and the production quality of the transistor is stable and balanced.

Description

GaN high electron mobility transistor big signal statistics model modelling approach
Technical field
The invention belongs to power device field, particularly to based on GaN high electron mobility transistor big Signal equivalent circuit statistical model modeling method.
Background technology
Gallium nitride (GaN) HEMT (HEMT, High Electron Mobility Transistor) due to its characteristic such as high frequency, high power, the application in microwave circuit is increasingly extensive.Due to Probabilistic impact such as technological fluctuation, needs statistical model to divide during GaN monolithic design Analysing its circuit performance, the biggest signal statistics model is the important step during organs weight.
Owing to GaNHEMT device has a biggest contact resistance, therefore first generation quasiconductor (silicon) and The big signal statistics model of second filial generation quasiconductor (GaAs, indium phosphide etc.) device can not directly apply to GaNHEMT device.The parameter extracting method of current main flow have employed PURVJANCE.J et al. more and carries Modeling technique [PURVJANCE.J, PETZOLD.M, the and POTRATZ.C, " A linear gone out statistical FET model using principal component analysis,”IEEE Trans.Microwave Theory Tech., vol.37, no.9, pp.1389-1394, Sep.1989], such as MEEHAN.M.D, The paper that WANDINGER.T, and FISHER.D.A et al. delivers: " Accurate design centering and yield prediction using the‘truth model’,”IEEE MTT-S International Microwave Symposium Digest, July, 1991, vol.1, pp.1201-1204. and J.F. The paper that Swidzinski et al. delivers: J.F.Swidzinski, K.Chang, " Nonlinear statistical modeling and yield estimation technique for use inMonte Carlo simulations,”IEEE Trans.Microwave Theory and Techniques,vol.48,no.12,pp.2316-2324,Dec. 2000。
The big signal statistics model modelling approach of main flow is to utilize DSMC to rebuild average, mark at present Accurate poor, correlation coefficient, simulates the large signal characteristic of multiple GaN HEMT device.But owing to covering spy Caro method is a kind of random algorithm, have employed whole large signal equivalent circuit parameters, and data volume is very big, The workload processing data is big.The problem not restrained can be run into, the mark to simulation result when data volume is the biggest The accuracy of quasi-difference has considerable influence.It is unfavorable for that the high efficiency to high-volume device models.
I.Angelov et al. proposes a kind of small signal equivalent circuit model modeling method [I.Angelov, M. Ferndahl,M.Gavell,G.Avolio,D.Schreurs,“Experiment design for quick statistical FET large signal model extraction,”Microwave Measurement Conference (ARFTG),81stARFTG,Seattle,WA,USA.5-8,Jun,2003,pp.1-5.].But the method is only Consider the situation that the sensitive parameter of a device changes simultaneously, the technique that different components can not be accurately reflected Situation of change.
Chen Zhikai et al. proposes a kind of big signal statistics model modelling approach based on DSMC [Zhikai Chen,Yuehang Xu,Bin Zhang,Tangsheng Chen,Tao Gao,RuiminXu,“A GaN HEMTs Nonlinear Large-Signal Statistical Model and Its Application in S-band PowerAmplifier Design,”IEEE Microw.Wirel.Compon.Lett,vol.26,no. 2,pp.128-130,Feb.2016].But the process employs DSMC to large signal equivalent circuit Parameter carries out statistical analysis, owing to the data volume of the method is very big, exceptional value easily occurs, therefore can cause The problem not restrained.
Summary of the invention
It is an object of the invention to provide a kind of GaN high electron mobility transistor big signal statistics model to build Mould method, this modeling method can reduce data volume, it is to avoid exceptional value occurs and runs into the problem that do not restrains, Enable the statistical model obtained by this modeling method to accurately reflect the technique change situation of different components, make The quality of production of transistor is stable, equilibrium.
For achieving the above object, the invention provides a kind of GaN high electron mobility transistor big signal system Meter model modelling approach, including:
Test several GaN high electron mobility transistors in each batch, it is thus achieved that described gallium nitride height electricity The I-E characteristic of transport factor transistor；
Described I-E characteristic is used for and the big signal etc. of described GaN high electron mobility transistor Effect circuit model, extracts the large signal equivalent circuit about described GaN high electron mobility transistor and joins Number；
According to described large signal equivalent circuit parameter, response-curve method is utilized to set up big signal statistics model.
Optionally, several GaN high electron mobility transistors in each batch of described test, it is thus achieved that institute The I-E characteristic stating GaN high electron mobility transistor includes:
Utilize probe station to test equipment, bias voltage is set, it is thus achieved that the drain-source electricity corresponding with described bias voltage Stream；
What described I-E characteristic referred to described bias voltage arranges numerical value and described drain-source current numerical value Corresponding relation.
Optionally, described bias voltage includes grid voltage Vgs and drain voltage Vds, described grid voltage The value of Vgs is from pinch-off voltage to 0V, and the value of described drain voltage Vds is for puncture to 1/2 from 0V Voltage.
Optionally, the method for described extraction large signal equivalent circuit parameter includes:
Utilize described large signal equivalent circuit model, use method of least square, come matching institute by matlab State I-E characteristic, it is thus achieved that the described big signal equivalence extracted from described large signal equivalent circuit model Circuit parameter.
Optionally, the described method utilizing response-curve method to set up big signal statistics model includes:
Choose several sensitive parameters in described large signal equivalent circuit parameter, define in described sensitive parameter Big sensitive parameter is XH, minimum sensitive parameter is XL, other described sensitive parameters are middle sensitive ginseng Number X, the computing formula of described middle sensitive parameter X is: X=b × C+a, wherein, a=(XH+XL)/2, B=(XH-XL)/2, the excursion of variable C is-1～1；
Choose three kinds of states of size variation-10%, 0 and 10% of each described sensitive parameter value, by some Described three kinds of states of individual described sensitive parameter are combined, and obtain N group analog parameter；
Described analog parameter is substituted in ADS software and emulates, it is thus achieved that big signal statistics model.
Optionally, described utilize response-curve method to set up big signal statistics model after, also include: checking is described The accuracy of big signal statistics model.
Optionally, the method for the accuracy of the described big signal statistics model of described checking includes:
Obtain the large signal characteristic of described GaN high electron mobility transistor；
Several sensitive parameters choosing described large signal equivalent circuit parameter substitute into described big signal statistics mould Type, obtains big signal simulation result；
Relatively described big signal simulation result and described large signal characteristic.
Optionally, described big signal simulation result includes simulation data power, simulated power added efficiency, imitates True gain.
Optionally, described large signal characteristic includes output Pout, power added efficiency PAE, gain Gain。
The specific embodiment provided according to the present invention, the invention discloses techniques below effect:
By testing several gallium nitride height in each batch in the statistical model modeling method that the present invention provides Electron mobility transistor obtains the IV characteristic of multiple GaN high electron mobility transistor, and often The IV characteristic variations of individual GaN high electron mobility transistor is separate, and by the change of each parameter Change situation is combined, and which includes parameter and increases compound mode when reducing with parameter, makes the present invention carry The statistical model that the modeling method of the big signal statistics model of confession obtains more can accurately reflect different components Technique change situation；Further, the statistical model modeling method of the present invention uses Response Surface Method, it is possible to according to Actual process range of error arranges parameter variation range, decreases the extracted amount of parameter value, more can effectively control The scope of error change, it is to avoid produce difference value, solves not restraining of traditional DSMC existence Problem so as to get model parameter value there is correct physical significance, it is possible to accurately reflect the work shape of device State.Meanwhile, this modeling method is easier and has efficientibility.The statistical model modeling method of the present invention Can be used for different large signal equivalent circuit models, it is possible to meet engineers and technicians and revise according to actual needs Large-signal model but do not change the requirement of statistical model modeling method；The statistics mould obtained by this modeling method Type can accurately reflect the technique change situation of different components, makes the quality of production of transistor stablize, equalize.
Accompanying drawing explanation
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, below will be to enforcement In example, the required accompanying drawing used is briefly described, it should be apparent that, the accompanying drawing in describing below is only Some embodiments of the present invention, for those of ordinary skill in the art, are not paying creative work On the premise of, it is also possible to other accompanying drawing is obtained according to these accompanying drawings.
The flow chart of the embodiment 1 of the big signal statistics model modelling approach that Fig. 1 provides for the present invention；
The flow chart of the embodiment 2 of the big signal statistics model modelling approach that Fig. 2 provides for the present invention；
The flow chart of the embodiment 3 of the big signal statistics model modelling approach that Fig. 3 provides for the present invention；
In the big signal statistics modelling verification accuracy step that Fig. 4 provides for the present invention, signal statistics model is imitated True result and the large signal characteristic comparison diagram of measured result；
Fig. 5 is that the present invention is calculated actual measurement and the average of Pout and PAE of emulation, the contrast of standard deviation Figure.
Detailed description of the invention
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clearly Chu, be fully described by, it is clear that described embodiment be only a part of embodiment of the present invention rather than Whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art are not making creation The every other embodiment obtained under property work premise, broadly falls into the scope of protection of the invention.
Embodiment 1: as described in Figure 1, the invention provides a kind of GaN high electron mobility transistor big Signal statistics model modelling approach, including:
Step 101: test several GaN high electron mobility transistors in each batch, it is thus achieved that nitridation The I-E characteristic of gallium HEMT；
Step 102: I-E characteristic is used for the big signal etc. with GaN high electron mobility transistor Effect circuit model, extracts the large signal equivalent circuit parameter about GaN high electron mobility transistor；
Step 103: according to large signal equivalent circuit parameter, utilizes response-curve method to set up big signal statistics mould Type.
The statistical model modeling method that present embodiment provides is by each batch in test different batches Several GaN high electron mobility transistors obtain multiple GaN high electron mobility transistor IV characteristic, and the IV characteristic variations of each GaN high electron mobility transistor is separate, And the situation of change of each parameter is combined, which includes parameter and increase combination side when reducing with parameter Formula, the statistical model making the modeling method of the big signal statistics model that the present invention provides obtain more can accurately be anti- Reflect the technique change situation of different components；Further, the statistical model modeling method of the present invention uses response surface design Method, it is possible to arrange parameter variation range according to actual process range of error, decreases the extracted amount of parameter value, More can effectively control the scope of error change, it is to avoid produce difference value, solve traditional DSMC Exist not convergence problem so as to get model parameter value there is correct physical significance, it is possible to accurately reflect The duty of device.Meanwhile, this modeling method is easier and has efficientibility.The statistics of the present invention Model modelling approach can be used for different large signal equivalent circuit models, it is possible to meet engineers and technicians according to It is actually needed amendment large-signal model but does not change the requirement of statistical model modeling method；By this modeling method The statistical model obtained can accurately reflect the technique change situation of different components, makes the quality of production of transistor Stable, equilibrium.
Unlike the embodiments above, as the optional embodiment of one, test several nitrogen in each batch Change gallium HEMT, it is thus achieved that the I-E characteristic of GaN high electron mobility transistor Including:
Utilize probe station to test equipment, bias voltage is set, it is thus achieved that the drain-source electricity corresponding with described bias voltage Stream；
What described I-E characteristic referred to described bias voltage arranges numerical value and described drain-source current numerical value Corresponding relation.
In the present embodiment, bias voltage includes grid voltage Vgs and drain voltage Vds, grid voltage The value of Vgs is from pinch-off voltage to 0V, and the value of drain voltage Vds is from 0V to 1/2 breakdown voltage. Obtain the current value between a drain electrode and the source electrode of correspondence by arranging two magnitudes of voltage, thus obtain nitrogen Change the I-E characteristic of gallium HEMT.This test mode can comprehensively and accurately reflect The I-E characteristic of GaN high electron mobility transistor, establishes accurate base for follow-up modeling process Plinth.
As the optional embodiment of one, the method extracting large signal equivalent circuit parameter includes:
Utilize large signal equivalent circuit model, use method of least square, carry out electricity described in matching by matlab Stream-voltage characteristic, it is thus achieved that the large signal equivalent circuit parameter extracted from large signal equivalent circuit model.
Embodiment 2: as in figure 2 it is shown, as a kind of optional embodiment, utilize response-curve method to set up The method of big signal statistics model includes:
Step 201: choose several sensitive parameters in large signal equivalent circuit parameter, in definition sensitive parameter Maximum sensitive parameter is XH, minimum sensitive parameter is XL, other sensitive parameters are middle sensitive parameter X, the computing formula of middle sensitive parameter X is: X=b × C+a, wherein, a=(XH+XL)/2, B=(XH-XL)/2, the excursion of variable C is-1～1；
Step 202: choose three kinds of shapes of size variation-10%, 0 and 10% of each described sensitive parameter value Described three kinds of states of several described sensitive parameters are combined, obtain N group analog parameter by state；
Step 203: described analog parameter is substituted in ADS software and emulates, it is thus achieved that big signal statistics model.
Owing to, in actual process, typical parameter value variation is about ± 10%, so choosing each parameter Three kinds of variable condition :-10%, 0,10%, the three state of several parameters is combined, permissible Obtain N group parameter, be used for simulating the parameter value of the statistical model of N number of pipe, so obtained model The situation of change of each parameter during considering actual process, more meets device technology change reality.And should Modeling method decreases the extracted amount of parameter value, the most simple and efficient, more can effectively control the model of error change Enclose, it is to avoid produce difference value, solve not convergence problem that traditional DSMC exists so as to get Model parameter value there is correct physical significance, it is possible to accurately reflect the duty of device.
Embodiment 3: as the optional embodiment of one, utilizes response-curve method to set up big signal statistics mould After type, also include: verify the accuracy of big signal statistics model.
In the present embodiment, verify that the method for the accuracy of big signal statistics model includes:
Step 301: obtain the large signal characteristic of GaN high electron mobility transistor；
Step 302: several sensitive parameters choosing large signal equivalent circuit parameter substitute into big signal statistics mould Type, obtains big signal simulation result；
Step 303: bigger signal simulation result and large signal characteristic.
In the present embodiment, big signal simulation result include simulation data power, simulated power added efficiency, Simulated gain.Large signal characteristic includes output Pout, power added efficiency PAE, gain G ain.
The invention provides a kind of efficient GaN HEMT (GaN high electron mobility transistor) big Signal statistics model modelling approach, enables GaN HEMT big signal statistics model reflecting device more accurately The electrology characteristic of part and microwave property, and make algorithm can apply to the device of different structure.
Below in conjunction with specific example, the modeling method of the present invention it is discussed in detail:
Step 1: test obtains the big signal code-voltage from 10 batch totally 34 GaN HEMT Characteristic:
Test obtain I-E characteristic: the biased electrical pressure point arranging each pipe includes: Vgs from-4V to 1V stepping 0.5V, totally 11 points, Vds is stepping 5V from 0 to 35V, totally 8 points；Obtain each partially Put the drain-source current value that electrical voltage point is corresponding, and then obtain I-E characteristic；
Step 2: I-E characteristic is used for the large signal equivalent circuit model of GaN HEMT, carries out Parameter extraction:
The Angelov GaN HEMT large-signal model of improvement used in the present invention is as GaN HEMT Large signal equivalent circuit model, its nonlinear source leakage current and gate capacitance model equation be:
Ids=Ipkth(1+Mipktanh(Ψ))tanh(αVds)
Ψ=P1th(Vgseff-Vpk1)+P2th(Vgseff-Vpk2)2+P3th(Vgseff-Vpk3)3
P i = ( P i 0 + P i 1 V d s ) tanh ( α P i ) + P i o , i = 1 , 2 , 3
Mipk=1+0.5 (Mipkbth-1)(1+tanh(Qm(Vgseff-Vgsm)))
Mipkb=(PM0+PM1Vds+PM2Vds 2+PM3Vds 3)tanh(αMVds)+PMo
QM=(PQ0+PQ1Vds)tanh(αQVds)+PQo
Vgseff=Vgs+γsurf1(Vgsq-Vgsqpinch)(Vgs-Vgsqpinch)+γsubs1(Vdsq+Vdssubs0)(Vds-Vdsq)
Ipkth=Ipk(T0)(1+KIpkΔT)
Pnth=Pn(T0)(1+KPnΔT)
MIpkbth=MIpkb(T0)(1+KMipkbΔT)
KIpk=(KIpk0+KIpk1Vds)tanh(αKIpkVds)+KIpko
KPn=(KPn0+KPn1Vds)tanh(αKPnVds)+KPno
KMipkb=(KKMipkb0+KMipkb1Vds)tanh(αKMipkbVds)+KMipkbo
Δ T=PdissRtheq=IdsVdsRtheq
Cgs=Cgsp+Cgs0×(1+tanh(φ1))×(1+tanh(φ2))
φ1=P10+P11×Vgs+P12×Vds
φ2=P20+P21×Vgs+P22×Vgs 2+P23×Vgs 3
Cgd=Cgdp+Cgd0×((1-P111+ tanh (φ3)) × (1+tanh (φ4)+2 × P111)
φ3=P30-P31×Vds
φ4=P40+P41×Vgs-P111×Vds
Utilize above-mentioned model equation, use method of least square, carry out matching I-E characteristic by matlab, Obtain the large signal equivalent circuit parameter extracted from large signal equivalent circuit model；
Step 3: use Response Surface Method to set up statistical model large signal equivalent circuit parameter:
Choose 4 parameters I most sensitive in large signal equivalent circuitPK, VPK1, Cgs0, Cgd0, it is made It is modeled by the second-order model of Response Surface Method:
y = β 0 + Σ i = 1 k β i x i + Σ i = 1 , j ≤ i k β i j x i x j - - - ( 21 )
Wherein, y is the summation of sensitive parameter value；K is the parameter value at each point；β0It it is constant value； βiIt it is linear coefficient；βijIt it is interaction coefficient；
In definition sensitive parameter, maximum sensitive parameter is XH, minimum sensitive parameter is XL, other are sensitive Parameter is middle sensitive parameter X, and the computing formula of middle sensitive parameter X is: X=b × C+a, wherein, A=(XH+XL)/2, b=(XH-XL)/2, the excursion of variable C is-1～1；
Owing to, in actual process, typical parameter value variation is about ± 10%, so choosing each parameter Three kinds of variable condition :-10%, 0,10%.The three state of 4 parameters is combined, can obtain To 34=81 groups of parameters, are used for simulating the parameter value of the statistical model of 81 pipes.
Step 4: the parameter value of the statistical model of 81 pipes is substituted in the ADS software of Agilent company Emulation obtains big signal statistics model.
Step 5: the biased electrical pressure point arranging each pipe includes: Vgs from-4V to 1V stepping 0.5V, Totally 11 points, Vds is stepping 5V from 0 to 35V, totally 8 points；Frequency from 0.1G to 40G, stepping 0.1G, totally 400 points, the bias point of large signal characteristic: Vgs=-2.8V, Vds=28V, frequency 3GHz； Thus obtain the large signal characteristic of GaN high electron mobility transistor: output Pout, power added Efficiency PAE, gain G ain；
Several sensitive parameters choosing large signal equivalent circuit parameter substitute into big signal statistics model, obtain big Signal simulation result: simulation data power, simulated power added efficiency, simulated gain；
Bigger signal simulation result and large signal characteristic, compare with the large signal characteristic of measured result and test Card, as shown in Figure 4, simulation result with measured result closely, and contains the scope of measured result； It is calculated actual measurement and the average of Pout and PAE of emulation, standard deviation, as it is shown in figure 5, simulation result With measured result the most closely, thus demonstrate the accuracy of model.
In this specification, each embodiment uses the mode gone forward one by one to describe, and what each embodiment stressed is With the difference of other embodiments, between each embodiment, identical similar portion sees mutually.
Principle and the embodiment of the present invention are set forth by specific case used herein, above enforcement The explanation of example is only intended to help to understand method and the core concept thereof of the present invention；Simultaneously for this area Those skilled in the art, according to the thought of the present invention, the most all can change Part.In sum, this specification content should not be construed as limitation of the present invention.

Claims (9)

1. a GaN high electron mobility transistor big signal statistics model modelling approach, its feature exists In, including:
Test several GaN high electron mobility transistors in each batch, it is thus achieved that described gallium nitride The I-E characteristic of HEMT；
Described I-E characteristic is used for the big signal of described GaN high electron mobility transistor Equivalent-circuit model, extracts the big signal equivalent electric about described GaN high electron mobility transistor LUSHEN number；
According to described large signal equivalent circuit parameter, response-curve method is utilized to set up big signal statistics model.
Modeling method the most according to claim 1, it is characterised in that in each batch of described test Several GaN high electron mobility transistors, it is thus achieved that the electricity of described GaN high electron mobility transistor Stream-voltage characteristic includes:
Utilize probe station to test equipment, bias voltage is set, it is thus achieved that the leakage corresponding with described bias voltage Source electric current；
What described I-E characteristic referred to described bias voltage arranges numerical value and described drain-source current number The corresponding relation of value.
Modeling method the most according to claim 2, it is characterised in that described bias voltage includes grid Pole tension Vgs and drain voltage Vds, the value of described grid voltage Vgs is from pinch-off voltage to 0V, The value of described drain voltage Vds is from 0V to 1/2 breakdown voltage.
Modeling method the most according to claim 1, it is characterised in that described extraction big signal equivalence The method of circuit parameter includes:
Utilize described large signal equivalent circuit model, use method of least square, carry out matching by matlab The curve of described I-E characteristic, it is thus achieved that described in extracting from described large signal equivalent circuit model Large signal equivalent circuit parameter.
Modeling method the most according to claim 1, it is characterised in that described utilize response-curve method The method setting up big signal statistics model includes:
Choose several sensitive parameters in described large signal equivalent circuit parameter, define described sensitive parameter The sensitive parameter of middle maximum is XH, minimum sensitive parameter is XL, during other described sensitive parameters are Between sensitive parameter X, the computing formula of described middle sensitive parameter X is: X=b × C+a, wherein, A=(XH+XL)/2, b=(XH-XL)/2, the excursion of variable C is-1～1；
Choose three kinds of states of size variation-10%, 0 and 10% of each described sensitive parameter value, will Described three kinds of states of several described sensitive parameters are combined, and obtain N group analog parameter；
Described analog parameter is substituted in ADS software and emulates, it is thus achieved that big signal statistics model.
Modeling method the most according to claim 1, it is characterised in that described utilize response-curve method After setting up big signal statistics model, also include: verify the accuracy of described big signal statistics model.
Modeling method the most according to claim 6, it is characterised in that the described big signal of described checking The method of the accuracy of statistical model includes:
Obtain the large signal characteristic of described GaN high electron mobility transistor；
Several sensitive parameters choosing described large signal equivalent circuit parameter substitute into described big signal statistics Model, obtains big signal simulation result；
Relatively described big signal simulation result and described large signal characteristic.
Modeling method the most according to claim 7, it is characterised in that described big signal simulation result Including simulation data power, simulated power added efficiency, simulated gain.
9. according to the modeling method described in any one of claim 1-8, it is characterised in that described big signal Characteristic includes output Pout, power added efficiency PAE, gain G ain.
CN201610203716.0A 2016-04-01 2016-04-01 Large signal statistical model modeling method for gallium nitride high electron mobility transistor Withdrawn CN105844042A (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201610203716.0A CN105844042A (en) 2016-04-01 2016-04-01 Large signal statistical model modeling method for gallium nitride high electron mobility transistor
CN201710120505.5A CN106845025B (en) 2016-04-01 2017-03-02 The big signal statistics model modelling approach of GaN high electron mobility transistor

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610203716.0A CN105844042A (en) 2016-04-01 2016-04-01 Large signal statistical model modeling method for gallium nitride high electron mobility transistor

Publications (1)

Publication Number Publication Date
CN105844042A true CN105844042A (en) 2016-08-10

Family

ID=56596663

Family Applications (2)

Application Number Title Priority Date Filing Date
CN201610203716.0A Withdrawn CN105844042A (en) 2016-04-01 2016-04-01 Large signal statistical model modeling method for gallium nitride high electron mobility transistor
CN201710120505.5A Expired - Fee Related CN106845025B (en) 2016-04-01 2017-03-02 The big signal statistics model modelling approach of GaN high electron mobility transistor

Family Applications After (1)

Application Number Title Priority Date Filing Date
CN201710120505.5A Expired - Fee Related CN106845025B (en) 2016-04-01 2017-03-02 The big signal statistics model modelling approach of GaN high electron mobility transistor

Country Status (1)

Country Link
CN (2) CN105844042A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107783022A (en) * 2017-10-27 2018-03-09 西安电子科技大学 The thermal reliability appraisal procedure of HEMT
CN110768517A (en) * 2019-11-20 2020-02-07 南方电网科学研究院有限责任公司 Control method for inhibiting field effect transistor switch oscillation and second-order model thereof

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107194127A (en) * 2017-06-30 2017-09-22 苏州芯智瑞微电子有限公司 A kind of transistor modeling method based on burst pulse weak-signal measurement
CN108520084B (en) * 2018-01-31 2020-04-17 电子科技大学 Microwave gallium nitride device nonlinear current model parameter extraction method and system
CN111144059A (en) * 2019-09-04 2020-05-12 苏州芯智瑞微电子有限公司 Simplification method and application of empirical capacitance model of gallium nitride high electron mobility transistor under high drain-source voltage condition
CN111428437B (en) * 2020-03-06 2023-05-02 电子科技大学 Method for extracting parameters of quasi-physical base statistical model of microwave transistor
CN112257367A (en) * 2020-11-11 2021-01-22 天津大学 Transistor small signal scalable model modeling method based on hybrid scaling rule
CN115618785B (en) * 2022-12-16 2023-03-21 电子科技大学 Harmonic mean function-based gallium nitride transistor physical fundamental large signal model

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102565650B (en) * 2010-12-07 2013-12-25 中国科学院微电子研究所 Measuring system of GaN HEMT device transconductance frequency scattering characteristic and method thereof
CN103913690A (en) * 2014-04-04 2014-07-09 中国科学院微电子研究所 Method and system for measuring frequency dispersion characteristic of transistor output resistance
CN104573330B (en) * 2014-12-20 2017-06-09 电子科技大学 The extracting method of GaN high electron mobility transistor I V model parameters

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107783022A (en) * 2017-10-27 2018-03-09 西安电子科技大学 The thermal reliability appraisal procedure of HEMT
CN110768517A (en) * 2019-11-20 2020-02-07 南方电网科学研究院有限责任公司 Control method for inhibiting field effect transistor switch oscillation and second-order model thereof
CN110768517B (en) * 2019-11-20 2021-06-29 南方电网科学研究院有限责任公司 Control method for inhibiting field effect transistor switch oscillation and second-order model thereof

Also Published As

Publication number Publication date
CN106845025B (en) 2019-11-05
CN106845025A (en) 2017-06-13

Similar Documents

Publication Publication Date Title
CN105844042A (en) Large signal statistical model modeling method for gallium nitride high electron mobility transistor
Liu et al. A Wiener-type dynamic neural network approach to the modeling of nonlinear microwave devices
CN106372357A (en) Method for establishing nonlinear noise model of GaN HEMT
CN107636656A (en) GaN device technological parameter statistical analysis technique based on large signal equivalent circuit model
CN102262206A (en) Method for predicting negative bias temperature instability (NBTI) service life of pMOSFET (P-channel Metal Oxide Semiconductor Field Effect Transistor) device
CN101655882B (en) Modelling method based on worst condition of statistic model
CN109800464A (en) Method, storage medium and the computer equipment of modeling and the design of semiconductor devices
CN102968528B (en) Numerical value simulation data processing method and device of semiconductor device
Jarndal et al. Efficient modeling of GaN HEMTs for linear and nonlinear circuits design
CN105226054B (en) A kind of general mismatch model and its extracting method
CN106909741A (en) A kind of modeling method of microwave GaN power devices
CN104331580A (en) Method for utilizing high-pressure field effect transistor sub circuit model to describe self-heating effect
US20130054219A1 (en) Equivalent Electrical Model of SOI FET of Body Leading-Out Structure, and Modeling Method Thereof
Zhu et al. A small-signal model extraction and optimization method for AlGaN/GaN HEMT up to 110 GHz
Jia et al. A threshold voltage model for charge trapping effect of AlGaN/GaN HEMTs
Zhang et al. Enabling automatic model generation of RF components: a practical application of neural networks
Pelaez-Perez et al. Experimental verification of analytical design equations based on X-parameters® for predicting role of series feedback
Jarndal Combined genetic algorithm and neural network technique for transistor modeling
Cao et al. Application of the genetic algorithm in MOSFET modeling and parameter extraction
Yuan et al. A novel hybrid parameter extraction method for GaAs/GaN HEMT modeling with electromagnetic analysis
Muff et al. From device modeling to characterization: a complete end to end design flow for SiC devices half bridge design
Marinković et al. Development of a neural approach for bias-dependent scalable small-signal equivalent circuit modeling of GaAs HEMTs
Ferranti et al. Effective electrothermal analysis of electronic devices and systems with parameterized macromodeling
Xu et al. Statistical models for microwave GaN HEMTs
CN104794318B (en) For establishing the data processing method of semiconductor devices statistical model

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
WW01 Invention patent application withdrawn after publication

Application publication date: 20160810

WW01 Invention patent application withdrawn after publication