Forschungszentrum Jülich in der Helmholtz-Gemeinschaft CWRF2008 RF Power improvement of AlGaN/GaN...
-
Upload
christopher-pereira -
Category
Documents
-
view
214 -
download
0
Transcript of 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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
21
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
27
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
eran
alys
is
32
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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 IDT=PT/VDS[15::1::40]
Eqn PT=1500
Eqn gm=diff(ID[::,20])
-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 IDT=PT/VDS[15::1::40]
Eqn PT=1500
Eqn gm=diff(ID[::,20])
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
CWRF2008
S-parameter analysis
freq (2.000GHz to 50.00GHz)
S(1
,1)
S(3
,3)
freq (2.000GHz to 50.00GHz)
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
freq (2.000GHz to 50.00GHz)
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
37
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
CWRF2008
Load pull simulation
…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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
CWRF2008
Load pull simulation
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
42
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
CWRF2008
Comparison of simulated and measured PAE
MOSHFET, SiO2 thickness:
10nm
Curves are acceptable between measurement and simulation of PAE
43
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
CWRF2008
Comparison of simulated and measured output power
MOSHFET, SiO2 thickness:
10nm
Good agreement between measurement and simulation of Pout
44
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
CWRF2008
CWRF2008
Thank you!
46
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
CWRF2008
ENDE Vortrag
47
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
CWRF2008
Load pull simulation
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
CWRF2008
Load pull simulation
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
CWRF2008
Load pull simulation varies the load reflection coefficient presented to a device...
Load pull simulation
…to find the optimal value to maximize output-power
51
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
CWRF2008
Modeling
From physical structure to equivalent circuit
52
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
CWRF2008
Modeling
Comparision of measured and simulated S-Parameter
S11
S12
S22
54
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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
Forschungszentrum Jülichin der Helmholtz-Gemeinschaft
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.