Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF2008 RF Power improvement of AlGaN/GaN...

Forschungszentrum Jülichin der Helmholtz-Gemeinschaft

CWRF2008

RF Power improvement of AlGaN/GaN based HFETs and MOSHFETs

RF Power improvement of AlGaN/GaN based HFETs and MOSHFETs

A. Fox, M. Marso

CWRF2008 CERN

2


CWRF2008

Research Center Juelich?

Where:

» between Cologne>>>>>Aachen

areas of research :

» M aterial

» E nvironment

» I nformation

» L ife

» E nergie

» Established more than 50 years

3


CWRF2008

Outline

Measurements

RF-power Simulation

Com

pari

son

of

sim

ulat

ed a

nd m

easu

red

RF

-pow

er p

erfo

rman

ce

Con

clus

ion

Mod

ellin

g S-Parameter

RF-powerLoad-Pull

Loa

d-P

ull

Sim

ulat

ion

DC

-, S

-par

amet

er

anal

ysis

introduction

4


CWRF2008

Application example

AlGaN/GaN HFETs are used in rf-generators, telecommunication and other applications where power devices at higher frequencies, voltages and temperatures are needed.

5


CWRF2008

AlGaN/GaN Heterostructure and 2DEG

This 2DEG “charge plate” creates the area with very high mobility and sheet concentration of carriers

Formation of 2DEG in AlGaN/GaN heterostructure

heterostructure is the layer system with different band gap material e.g. with AlGaN grown on GaN.

High Electron Mobility Transistor

D

Ohmic contact

SOhmic contact

GSchottky contact

Ohmic contacts are controlled by the Schottky Gate conntact

6


CWRF2008

How to improve power performance

Pout= Imax* (Vbreakdown-Vknee ) /8

for increasing Pout an output voltage- and current increase is needed

Example I-V characteristic

Imax increases due to increase carrier density and velocity

Vbreakdown increases e.g.

-with fieldplate technology-wide bandgap

ClassA operating point

Imax Max. output power: f(Ids,Vds)

Opt. RL=(Vb-Vs)/Imax

Vknee Vbreakdown

7


CWRF2008

Theory of HFETs: DC behavior

Drain Current (Ids)

How do the DC measures translate into device geometry and material properties?

ds

dsd d

VWqnI

qnvWId

Transconductance (gm)

h

WLG

qn

0rgs

GrG V

h

WLWqnL 0

qnvWId

dsrd vVh

WI 0

vh

W

V

Ig r

ds

dm 0

absolute charge underneath the gate equals charge in active region of 2DEG:

insertion of expression into saturated current formula...

...yields:

differentiation yields intrinsic transconductance

(robertson2001a)

the formulas tell us to:•charge carrier concentration •mobility, velocity

Id

Vd

8


CWRF2008

Advanced Material Properties

1.22.92.90.050.51.3, W/cm-K

1.5-2220.70.81vsat, 107cm/s

2.5222.5-321vpeak, 107cm/s

680340610700047007100, cm2/Vs

40303346.45.7EBR, 105 V/cm

3.433.20.71.41.1EG, eV

GaN6H SiC4H SiCInGaAs*GaAsSi

Important for high output power:- high breakdown field and voltage, i.e. wide bandgap-high thermal conductivity

Important for high fT and fmax : Fast carriers (i.e. µ0, vpeak , vsat)

Frank Schwierz Fachgebiet Festkörperelektronik, Technische Universität Ilmenau, Germany

9


CWRF2008

HFET performance improvement

*Juraj Bernat „Fabrication and characterisation of AlGaN HEMT“ (Diss `2005@RWTH Aachen)

Enlarging of the Source - Drain distance is limited

degradation of dc and transport properties due tolong S_D distance

10


CWRF2008

HFET performance improvement

Fieldplate technology: Known since 1969 on Si

Higher breakdown voltage Different electric field

distribution between Gate and Drain

*Juraj Bernat „Fabrication and characterisation of AlGaN HEMT“ (Diss `2005@RWTH Aachen)

Performance improved by Fieldplate technology on AlGaN-HEMT

11


CWRF2008

Fieldplate Technology - Results

Increase of the output power is due to modification of the electric field on GaN cap

Confirmed by: A. Koudymov, IEEE Electr. Device L. 26, Oct. 2005*Juraj Bernat „Fabrication and characterisation of AlGaN

HEMT“ (Diss `2005@RWTH Aachen)

Simulation: by ATLAS, Silvaco

12


CWRF2008

Experiment

Unpassiv. HFET

• Simultaneous fabrication of unpassivated & passivated SiC/GaN/AlGaN HFETs and MOSHFETs:

• 10 nm SiO2 layer for passivated HFETs and MOSHFETs (PECVD),

• LG=0.3-0.9 µm, LW =200 µm (2-fingers),

Source

Source

Drain

GateS D

SiO2 ~10 nm

passivated HFET

SiO2 ~10 nm

MOSHFET

Gero Heidelberger:Technology related issues regarding fabrication

4H-SiC

3 µm GaN

30 nm Al0.28Ga0.72N

G

13


CWRF2008

HFET- technology

• Mesa etching– ECR RIE or Ar+ sputter– Depth: 250 ~ 300 nm

• Ohmic contacts– Ti/Al/Ni/Au– Annealing: 850ºC, 30sec

• Schottky contacts– Ni/Au

• Pads– Ti/Au

Fabrication of our HFET Devices

14


CWRF2008

HFET Layout

SEM picture of AlGaN/GaN HFET wg=200 µmLg=0.3 µm

Hetero Field Effect Transistor

SEM picture of AlGaN/GaN HFET with airbridge technology >> increasing Imax

Detailed view of HFET Device as fabricated in our labprepared for on-wafer measurements with 100 µm pitch

15


CWRF2008

Outline

Measurements

RF-power Simulation

Com

pari

son

of

sim

ulat

ed a

nd m

easu

red

RF

-pow

er p

erfo

rman

ce

Mod

ellin

g

S-ParameterRF-powerLoad-Pull

Loa

d-P

ull

Sim

ulat

ion

DC

-, S

-par

amet

er

anal

ysis

16


CWRF2008

S-parameter Measurement System

• Important for microwave design and characterization

• Basics for parameter extraction and simulation

• Our measurement system:

– Frequency up to 110 GHz

– Network Analyzer

– On wafer measurements

– Control System and Calibration (LRM)

17


CWRF2008

110 GHz Measurement System

We are well prepared for high quality s-parameter measurements

HP8510XF NWA-System:-Network Analyser HP8510C - RF source 45 MHz 50GHz Synth. sweeper HP 83651 -LO source 45MHz to 20GHz Synth. Sweeper HP 83621 -mm-wave controller -Cascade probestation-Connectors: 1 mm coaxial-Coplanar Probetips 110 GHz-LRM Calibration , attanuation

DUT mm-wave controllermm-wave controller

FZ-Jülich, ISG1, A. Fox

18


CWRF2008

Results of S-parameter measurements

1E10 1E11-10

-5

0

5

10

15

20

25

freq.[Hz]

h21

^2[d

B]

3c HFET 3a MOSHFET 3d pas.HFET

cut off frequency HFET, MOSHFET

MOSHFET:ft=24GHz

HFET:ft=17GHz

drain source voltage: 20V gate length: 500 nmSiO2 layer thickness: 10nm

cut off frequency (ft) is the frequency where current gain h21=0 [db]

MOSHFETs show significantly higher cut off frequencies

19


CWRF2008

Variation of ft vs. gate length

Higher cut-off frequencies (ft) are observed for MOSHFETs for all kinds of gate lengths. ft increases with decreasing gate lengths.

20


CWRF2008

Outline

Measurements

RF-power Simulation

Com

pari

son

of

sim

ulat

ed a

nd m

easu

red

RF

-pow

er p

erfo

rman

ce

Mod

ellin

g


Loa

d-P

ull

Sim

ulat

ion

DC

-, S

-par

amet

er

anal

ysis

21


CWRF2008

RF-Power measurement

Load Pull Measurement System

step attenuator

Microwave Tuner Systemslotted coaxial linecomputer controlled slug Tuner cal.deembedingFocus Microwave

Ref.plane

resonanz.

22


CWRF2008

RF-Power measurement system

FZ-Jülich, ISG1, A. Fox

Large signal measurements were performed on wafer at 7 GHz by source and load pull measurements

Load-tuner

Source-tuner

NetworkanalyzerHP8510B

DUT

Bias supply

23


CWRF2008

10 15 20 250

4

8

12

16

20

24

28

32

0

10

20

30

40

50

Uds

= 12.7V

Uds

= 17.5V

Uds

= 22.4V

Pout ([dBm]) Gain ([dB])

Po

ut (

dB

m)

; G

ain

(d

B)

Pin (dBm)

@7GHz

P.A

.Eff

(%

)

P.A.Eff ([%])

Sample: S2661_3d M11_jDate: 14.02.2008

VG = -2.5 V

Power performance will increase with further increase of Vds for passivated HFET

Load pull measurement

24


CWRF2008

Power measurement at differend SiO2layer

Comparison of Pout for HFET and MOSHFET with different SiO2 dielectric layer

@ 7 GHz

0 5 10 15 20 2515

16

17

18

19

20

21

22

23

24

25

26

27

P

ou

t[d

Bm

]

Pin[dBm]

5nm SiO2

10nm SiO2

20nm SiO2

HFET

Bias:Ug=-3V,Ud=20V

Performance improvement by optimising the dielectric layer thickness

25


CWRF2008

Comparison of Power measurements

The RF output power increases from 2.9 W/mm (HFET) to 6.7 W/mm (MOSHFET) with no fieldplate implemented

SiO2 layer thickness: 10nm

WG = 200 µm , LG = 0.7 µm f = 7 GHzVg = -2.5 V HFETVg = -2.5 V passiv. HFETVg = -7 V MOSHFET

12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

2

3

4

5

6

7

HFET

MOSHFET

passivated HFET

WG = 200 m, L

G = 0,7 m

f = 7 GHzU

G = -2.5 V HFET

UG = -2.5 V passiv. HFET

UG = -7 V MOSHFET

Po

ut (

W/m

m)

Vds [V]

26


CWRF2008

Outline

Measurements

RF-power Simulation

Com

pari

son

of

sim

ulat

ed a

nd m

easu

red

RF

-pow

er p

erfo

rman

ce

Mod

ellin

g


Loa

d-P

ull

Sim

ulat

ion

DC

-, S

-par

amet

er

anal

ysis

27


CWRF2008

Cgs and gm change with the dielectric layer

Modelling

Equivalent circuit for the AlGaN/GaN MOSHFET

S

GD

S

ft =f( gm/Cgs)Oxydedeplition

28


CWRF2008

Extracted Parameters Cgs, gm

Cgs decreases for MOSHFETs due to additional SiO2 layer

vh

W

V

Ig r

ds

dm 0

vsat increases with passivation

From theory:

29


CWRF2008

Modelling

Comparison of gm/Cgs ratio

gm/Cgs is in agreement with the increase of cut off frequency from h21

for the MOSHFET compared to HFET

ft =f( gm/Cgs)

30


CWRF2008

Modelling

MOSHFETVg=-1VVds=19VFrequency: 2 to 50GHz

S22

S11

Good agreement between measurement and optimized model

Measurement: blue curvesModel: red curves

Comparison of measurements and modelS21

S12

Parameterextraction done by TOPAS, IMST

31


CWRF2008

Outline

Measurements

RF-power Simulation

Com

pari

son

of

sim

ulat

ed a

nd m

easu

red

RF

-pow

er p

erfo

rman

ce

Mod

ellin

g


Loa

d-P

ull

Sim

ulat

ion

DC

-, S

-par

amet

eran

alys

is

32


CWRF2008

DC analysis

VGSTj

VDS

TOPASTOPAS1

Conv_dy=0.001Conv_dx=0.001W=-1N=-1ShiftVp=0.0DeltaG=1.0DeltaRi=1.0DeltaIds=1.0DeltaCds=1.0DeltaCgd=1.0DeltaCgs=1.0Simtype=SplinesFile="C:\users2005\topas_22_8_2006_fox_prj\data\2_1076_1_3_L14k_2.sim"

Ti

Tj

VARVAR1

VD=3VG=-0.2

EqnVar

V_DCSRC_DVdc=VD V

V_DCSRC_GVdc=VG V

I_ProbeID

V_DCSRC_TiVdc=T V

I_ProbeIG

DC design circuit DC simulatorbias sweepDUT with the modeled parameters

Advanced Design System, Agilent

DUT

33


CWRF2008

Transfer I-V-Curves

2 4 6 8 10 12 14 16 180 20

0

50

100

150

-50

200

VDS [V]

ID [m

A] Transconductance

Output I-V-Curves

Eqn VDSindex=[20]

Input I-V-Curves

Eqn IDT=PT/VDS[15::1::40]

Eqn PT=1500

Eqn gm=diff(ID[::,20])

Transfer I-V-Curves

Transconductance

Output I-V-Curves

Eqn VDSindex=[20]

Input I-V-Curves


Eqn PT=1500


-4 -3 -2 -1 0-5 1

0

50

100

150

-50

200

VGS [V]

ID [m

A]

Transfer I-V-Curves

-4 -3 -2 -1 0-5 1

10

20

30

0

40

VGS [V]

gm [m

S]

Transconductance

Output I-V-Curves

Eqn VDSindex=[20]

Input I-V-Curves


Eqn PT=1500


DC analysis: Results

The curves are comparable to DC-measurements and reflect the intrinsic behavior

Output I-V- curves

Transfer I-V- curves

Transconductance

34


CWRF2008

S-parameter analysis

S-parameter design circuit

Tj

VDVG

TOPASTOPAS1

Conv_dy=0.001Conv_dx=0.001W=-1N=-1ShiftVp=0.0DeltaG=1.0DeltaRi=1.0DeltaIds=1.0DeltaCds=1.0DeltaCgd=1.0DeltaCgs=1.0Simtype=SplinesFile="C:\users2005\topas_22_8_2006_fox_prj\data\2_f34_O11i.sim"

Ti

Tj

VARVAR1

VD=16VG=-2

EqnVar

LL3

R=Rs OhmL=Ls nH

V_DCSRC_TiVdc=T V

V_DCSRC_DVdc=VD V

PortP2Num=2

I_ProbeID

DC_FeedDC_Feed2

DC_BlockDC_Block2

LL2

R=Rd OhmL=Ld nH

LL1

R=Rg OhmL=Lg nH

DC_BlockDC_Block1

DC_FeedDC_Feed1

VARVAR3

Rs=0Rd=0Rg=0

EqnVarVAR

VAR2

Ls=0Ld=0Lg=0

EqnVar

PortP1Num=1

I_ProbeIG

V_DCSRC_GVdc=VG V

S-parameter simulatorfrequency sweep

DUT

35


CWRF2008

S-parameter analysis

freq (2.000GHz to 50.00GHz)

S(1

,1)

S(3

,3)


S(2

,2)

S(4

,4)

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

-2.5

2.5


S(2

,1)

S(4

,3)

measured: red curvessimulated: blue curves

Comparison of Transistor Scattering Parameters

S11 S21

S22

different fitting at dif. biaspointsneeds further optimazation@that working point

36


CWRF2008

Outline

Measurements

RF-power Simulation

Com

pari

son

of

sim

ulat

ed a

nd m

easu

red

RF

-pow

er p

erfo

rman

ce

Mod

ellin

g


Loa

d-P

ull

Sim

ulat

ion

DC

-, S

-par

amet

er

anal

ysis

37


CWRF2008

• Specifying and generating desired load and source reflection coefficients (impedance)

• Biasing the device and running a simulation

• Calculating desired responses (delivered power, PAE, etc.)

Load-pull simulation

38


CWRF2008

Load pull simulation

Load pull simulation varies the load reflection coefficient presented to a device...

DUT

Tj

Bond wires:

Refer to the data display file "ReflectionCoefUtility"for help in setting s11_rho and s11_center.Also, refer to the example design file: examples/RF_Board/LoadPull_prj/HB1Tone_LoadPull_eqnsfor details about how this simulation is run.

s11_rho is the radiusand s11_center is thecenter of the circle.(But this is just astatic drawing.)

Specify desired Fundamental Load Tuner coverage:s11_rho is the radius of the circle of reflection coefficients generated. However, the radius of the circle will be reduced if it would otherwise go outside the Smith Chart.s11_center is the center of the circle of generated reflection coefficientspts is the total number of reflection coefficients generatedZ0 is the system reference impedance

Set Load and Source impedances at harmonic frequencies:

Set these values:

Vs_high

vload

Vs_low

Data Display Window:

DISP_HB_1Tone_LoadPull

One Tone Load Pull Simulation;output power and PAE found ateach fundamental load impedance

TOPASTOPAS1

Conv_dy=0.001Conv_dx=0.001W=-1N=-1ShiftVp=0.0DeltaG=1.0DeltaRi=1.0DeltaIds=1.0DeltaCds=1.0DeltaCgd=1.0DeltaCgs=1.0Simtype=SplinesFile="C:\users2005\topas_22_8_2006_fox_prj\data\2_fox.sim"

TiT j

OptionsOptions1

I_AbsTol=1e-8 AI_RelTol=1e-3V_AbsTol=1e-4 VV_RelTol=1e-3Tnom=TTemp=T

OPTIONS

VARSweepEquations

Z0=50pts=200s11_center=0.60+j*0.20s11_rho=0.3

EqnVar

HarmonicBalanceHB1

Order[1]=5Freq[1]=RFfreq

HARMONIC BALANCE

VARSTIMULUS

T=25Vlow=-1.5Vhigh=15RFfreq=7.5 GHzPavs=20 _dBm

EqnVar

ParamSweepSweep1

PARAMETER SWEEP

VARImpedanceEquations

EqnVar

VARBond_wires

Rs=0Rd=0Rg=0Ls=0Ld=0Lg=0

EqnVar

VARVAR2

Z_s_5=Z0+j*0Z_s_4=Z0+j*0Z_s_3=Z0+j*0Z_s_2=Z0+j*0Z_s_fund=50+j*0Z_l_5=Z0+j*0Z_l_4=Z0+j*0Z_l_3=Z0+j*0Z_l_2=Z0+j*0

EqnVar

VARVAR1

Z_s_5=10*Z0+j*0Z_s_4=10*Z0+j*0Z_s_3=10*Z0+j*0Z_s_2=10*Z0+j*0Z_s_fund=50+j*0Z_l_5=10*Z0+j*0Z_l_4=10*Z0+j*0Z_l_3=10*Z0+j*0Z_l_2=10*Z0+j*0

EqnVar

V_DCSRC_TiVdc=T V

LL4

R=Rd OhmL=Ld nH

CC2C=1 F

LL2

R=L=1 H

I_ProbeIs_high

V_DCSRC2Vdc=Vhigh

I_ProbeIload

S1P_EqnS1S[1,1]=LoadTunerZ[1]=Z0

P_1TonePORT1

Freq=RFfreqP=dbmtow(Pavs)Z=Z_sNum=1

V_DCSRC1Vdc=Vlow

I_ProbeIs_low

LL1

R=L=1 H

CC1C=1 F

LL3

R=Rg OhmL=Lg nH

LL5

R=Rs OhmL=Ls nH

Load- Pull design circuitHarmonic-Balance-simulatorfor nonlinear circuit behavior

?

ADS © Agilent

39


CWRF2008


…to find the optimal value to maximize power or PAE, etc.

This data shows the simulation results when the reflectioncoefficient is the value selected by the marker’s location.

40


CWRF2008


VinputTemperature

Vload

Vs_high

Vs_ low

Data Display Windows:

DISP_HB_2Tone_ topas_1

Harmonic Balance 2Tone Simulation; one input frequency; swept power

Includes PAE Calculation

TOPASTOPAS2

W=- 1N=- 1Simtype=SplinesFile="C:\users2005\topas_ fox_prj\data\2_ fox.sim"

Ti

Tj

LL4

R=Rd OhmL=Ld nH

DC_FeedDC_Feed4

DC_FeedDC_Feed3

DC_FeedDC_Feed2

DC_FeedDC_Feed1

V_DCSRC2Vdc=Vhigh V

I_P robeIs_high

TermTerm2

Z=Zl OhmNum=2

I_P robeIload

LL6

R=Rs OhmL=Ls nH

I_P robeIinput

LL3

R=Rg OhmL=Lg nH

V_DCSRC1Vdc=Vlow V

I_P robeIs_ low

P_nTonePORT1

P[2]=dbmtow(RFpower- 3)P [1]=dbmtow(RFpower- 3)Freq[2]=RFfreq+fspacing/2Freq[1]=RFfreq- fspacing/2Z=Zs OhmNum=1

Harmonic Balance simulator Input power sweep

Non linear circuit simulation results in Pout and PAE of the DUT

DUT

MatchedInput impedance

Matched Output impedance

41


CWRF2008

Outline

Measurements

RF-power Simulation

Com

pari

son

of

sim

ulat

ed a

nd m

easu

red

RF

-pow

er p

erfo

rman

ce

Mod

ellin

g


Loa

d-P

ull

Sim

ulat

ion

DC

-, S

-par

amet

er

anal

ysis

42


CWRF2008

Comparison of simulated and measured PAE

MOSHFET, SiO2 thickness:

10nm

Curves are acceptable between measurement and simulation of PAE

43


CWRF2008

Comparison of simulated and measured output power

MOSHFET, SiO2 thickness:

10nm

Good agreement between measurement and simulation of Pout

44


CWRF2008

Conclusion

• Increase of RF-performance by introduction of the MOSHFET-technology.

- MOS-HFET superior to unpassivated and passivated HFET regarding DC-, RF-, and power-performance.

• Increase of output power, PAE and cut-off frequency.

• Simulations verify the measured large signal results.

• Further work: alternative dielectric material and additional fieldplate implementation

45


CWRF2008

CWRF2008

Thank you!

46


CWRF2008

ENDE Vortrag

47


CWRF2008

RF-Power measurement (Load-Pull )

Comparison of RF-Power performance for passivated HFET and MOSHFET

Matched @max-PoutPAE=(Pout-Pin)/Pdc

MOSHFETHFET

MOSHFETs show significantly higher output Power and PAE

48


CWRF2008


matched loadimpedance

DUT

Harmonic BalanceInput power sweep

Load-Pull design circuit

matched source impedance

a1 b1 b2 a2

Vin Vout

topas_2pX1

OptionsOptions1

I_AbsTol=1e-8 AI_RelTol=1e-3V_AbsTol=1e-4 VV_RelTol=1e-3Tnom=TTemp=T

OPTIONS

HarmonicBalanceHB1

Other=OutVar="Z0" OutVar="Zs" OutVar="Zl"SweepPlan="Coarse1"SweepVar="RFpower"UseKrylov=noSS_Freq= SS_MixerMode=Order[1]=5Freq[1]=RFfreqMaxOrder=0

HARMONIC BALANCE

MeasEqnImpedance

LZ22=(1+LS22)/(1-LS22)*Z0[::,1]LZ11=(1+LS11)/(1-LS11)*Z0[::,1]

EqnMeas

VARVAR1

T=25RFpower=-1000RFfreq=7.50 GHz

EqnVar

TermTerm2

Z=Zl OhmNum=2

MeasEqnInput_impedance_and_matching

Gamma_DUT_in=10*log((dbmtow(RFpower)-dbmtow(Pin1))/dbmtow(RFpower))Z_DUT_in=HB1.HB.Vin[::,1]/HB1.HB.Iin.i[::,1]

EqnMeas

MeasEqnLSij

LS22=HB2.HB.b2[::,1]/HB2.HB.a2[::,1]LS12=HB2.HB.b1[::,1]/HB2.HB.a2[::,1]LS21=HB1.HB.b2[::,1]/HB1.HB.a1[::,1]LS11=HB1.HB.b1[::,1]/HB1.HB.a1[::,1]

EqnMeas

ParamSweepSweep1

Step=0.5 GHzStop=40 GHzStart=30 GHzSimInstanceName[6]=SimInstanceName[5]=SimInstanceName[4]=SimInstanceName[3]=SimInstanceName[2]=SimInstanceName[1]="HB1"SweepVar="RFfreq"

PARAMETER SWEEP

S4P_EqnSimpleCoupler2

S[4,2]=1S[3,1]=1S[2,1]=1S[1,2]=1

S4P_EqnSimpleCoupler1

S[4,2]=1S[3,1]=1S[2,1]=1S[1,2]=1 I_Probe

IinI_ProbeIout

TermTerm3

Z=Zs OhmNum=3

TermTerm4

Z=Zs OhmNum=4

TermTerm5

Z=Zl OhmNum=5

TermTerm6

Z=Zl OhmNum=6

P_1TonePORT1

Freq=RFfreqP=dbmtow(RFpower)Z=Zs OhmNum=1

Non linear circuit simulation results in Pout and PAE of the DUT

49


CWRF2008


Tj

VDVG

TOPASTOPAS1

Conv_dy=0.001Conv_dx=0.001W=-1N=-1ShiftVp=0.0DeltaG=1.0DeltaRi=1.0DeltaIds=1.0DeltaCds=1.0DeltaCgd=1.0DeltaCgs=1.0Simtype=SplinesFile="C:\users2005\topas_22_8_2006_fox_prj\data\2_f34_O11i.sim"

Ti

Tj

VARVAR1

VD=19VG=-1

EqnVar

LL3

R=Rs OhmL=Ls nH

V_DCSRC_TiVdc=T V

V_DCSRC_DVdc=VD V

PortP2Num=2

I_ProbeID

DC_FeedDC_Feed2

DC_BlockDC_Block2

LL2

R=Rd OhmL=Ld nH

LL1

R=Rg OhmL=Lg nH

DC_BlockDC_Block1

DC_FeedDC_Feed1

VARVAR3

Rs=0Rd=0Rg=0

EqnVarVAR

VAR2

Ls=0Ld=0Lg=0

EqnVar

PortP1Num=1

I_ProbeIG

V_DCSRC_GVdc=VG V

DUT

Detail circuit of Load-Pull design circuit

50


CWRF2008

Load pull simulation varies the load reflection coefficient presented to a device...


…to find the optimal value to maximize output-power

51


CWRF2008

Modeling

From physical structure to equivalent circuit

52


CWRF2008

CV-measurement

-13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 00

50

100

150

200

250

300 HFET passivated HFET MOSHFET

cap

acit

ance

(n

F/c

m2)

gate voltage (V)

53


CWRF2008

Modeling

Comparision of measured and simulated S-Parameter

S11

S12

S22

54


CWRF2008

Output Power

8

)(.

kneebrsatddc

linout

VVIP

2.

16

8

)(

kneebr

satddc

satout

VVIP

To reach maximum Pout:

– High Idsat

– High Vbr

– Small Vknee Small Rs

increased output power by increasing Id and Ud

55


CWRF2008

Reduction of Gate Leakage Current

MOSHFET using 11nm thick SiO2 gate isolation:• 30% increase of the drain saturation current• Reduction of gate leakage current from 10-6 to 10-10 A/mm

*Juraj Bernat „Fabrication and characterisation of AlGaN HEMT“

(Diss `2005@RWTH Aachen)

56


CWRF2008

Wide bandgap semiconductors

Common semiconductorsEG around 1 eVSi EG = 1.1 eVGaAs EG = 1.4 eV

E

EC

Ei

EV

Conduction band

Valence band

EGWide bandgap semiconductorsEG > 2 eV4H SiC EG = 3.2 eV6H SiC EG = 3.0 eVGaN EG = 3.4 eVAlN EG = 6.1 eVAlGaN EG = 3.4 ... 6.1 eVC EG = 5.5 eVNarrow bandgap semiconductorsEG << 1 eVInAs EG = 0.35 eVInSb EG = 0.17 eV

57


CWRF2008

SiC and GaN based Transistors

Cree, Inc : Pout= 32.2 W/mm, PAE= 54,8%, Fe_doped and Field plate GaN HEMT @10GHz Pout=4.1W/mm, PAE=72%

58


CWRF2008

Heterostructure and 2DEG

A heterostructure is the layer system where two semiconductors with different bandgaps Eg are grown one on the otherIn the thermodynamical equilibrium,when both semiconductors are "connected" together, the Fermi-level energy (EF ) of the Semiconductor I and Semiconductor II must be in the line what cause the discontinuity in the conductance (EC) and valence (EV ) band and the band bending. This results in the formation of the triangular quantum well where carriers are fixed in one axis and the 2DEG is formed.

Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF2008 RF Power improvement of AlGaN/GaN...

Documents

Transcript of Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF2008 RF Power improvement of AlGaN/GaN...