C²S²
WorkshopWMB
Nonlinear characterization and modeling of low
frequency dispersive effects infrequency dispersive effects in power transistors
R. Quéré(1), O. Jardel(2), A. Xiong(1), M. Oualli(2), T. Reveyrand(1),
J.P. Teyssier(1), R. Sommet(1), J.C Jacquet(2) , S. Piotrowicz(2)
(1) MITIC/XLIM University of Limoges(2) MITIC/Thales 3-5 Lab
-1-
C²S² Outline
Characterization methods
Th l IThermal Issues
Trapping effectsTrapping effects
Impact Ionization effectsImpact Ionization effects
ConclusionConclusion
-2- 19/12/2008
C²S² Challenges for the design of HPA
Design of High Power Amplifiers requires accurate Non linear models that take into accurate Non linear models that take into account:
Strong Thermal constraintsStrong Thermal constraintsParasitics effects such as Traps in HEMTsImpact Ionization limits in GaAs PHEMts
To cope withTo cope withReliability issuesD d ti f L i l h t i tiDegradation of Large signal characteristics
Rely on specialized characterizations tools
-3-
y p
19/12/2008
C²S²
Characterization ToolsCharacterization Tools
-4-
C²S² Characterization of microwave devices
In the modelling process, the characterization phase is preeminent, as some effects can only be put into
id i i li d h ievidence using specialized measurements techniques.
DC and Pulsed I-V measurements
CW and pulsed S-parameters meaasurements
CW d l d L d P ll f tCW and pulsed Load-Pull frequency measurements
CW and pulsed Load-Pull time domain measurements
(LSNA)
Low Frequency Z and S parameters measurementsLow Frequency Z and S parameters measurements
Two tone IM measurements.
-5-
C²S² Pulsed Measurements of power devices
g µ
Vgs=+1.000 VVgs= +0.00nVVgs=-1.000 VVgs=-2.000 VVgs=-3.000 V
• Max current 5A
On WaferOn Wafer performances performances Vgs=-4.000 VVgs=-5.000 VVgs=-6.000 VVgs=-7.000 VVgs=-8.000 VVgs=+1.000 VVgs= +0.00nVVgs=-1.000 VVgs=-2.000 V
• Min current 1µA• Max Voltage 120V• S-parameters 1- 40 GHz
Vgs=-3.000 VVgs=-4.000 VVgs=-5.000 VVgs=-6.000 VVgs=-7.000 VVgs=-8.000 V
•Température -65°C / 200°C • Pulse duration > 300ns • duty cycle > 0.5 %
-6- 6
C²S²Principle of pulsed measurements
- short pulses 400ns : quasi –isothermal state
- period : 6µs
- starting point of the pulses is the quiescent bias point (Vgs0,Vds0) that defines the thermal and trapping state of e a a d app g s a e othe device
- small-signal RF during the steady state of the pulses
- I(V) and S-Parameters are taken during the pulsestaken during the pulses
-7-
C²S² Harmonic Load Pull ( Frequency domain analysis)
RF coupler coupler
50Ω
p p
DC biasLin Amp
D U T HN Tuner DC bias
10 MhzSynch 4 channel test set y
LO f BP
VNA receiver operation mode : Heterodyne Principle + narrow BP IF filter
LO frequency synthesis
BPFilter
Detection
Mixer
Sequential Measurements of Harmonic components Hn No phase relationship measurements between Hn and Hn+1
VNA receiver operation mode : Heterodyne Principle + narrow BP IF filter
-8-
Absolute power and power wave ratios measurements @ Hn
C²S² Load pull Time Domain Waveform Measurements
50Ω
ReferenceComb generatorFor phase cal procedure
RF coupler coupler
50Ω
Lin AmpD U T
10 Mh
Lin AmpDC bias DC biasHN Tuner
Time equivalent Waveforms @IF
10 MhzSy nch
10 20 Mhz Comb li 10 Mh 20 Mh
Harmonic Sub Sampling @ 20 MHz
10-20 MhzSynthesizer
CombGen
samplinghead
10 MhzLP filter
20 MhzADC
-9-
Frequency Translation and compression into a 10 Mhz IF Bandwidth
C²S² Pulsed LSNA-organisation
-10-
C²S² Pulsed LSNA- The stroboscopic approach
-11-
C²S² Pulsed LSNA- The set-up
-12-
C²S²
Thermal issuesThermal issues
-13-
C²S² Thermal Behaviour Impact on PA
Gain Average current
Degradation of performances for a 10W X band PA
O t t P & PAE Junction Temperature
100
120
140
era
ture
(°C
)
Outpout Power & PAE Junction Temperature
60
80
40Jun
ctio
n te
mp
e0 5 10 15 20-5 25
40
Pin (dBm)
J
Optimization of PAE allows to reach the relialbility
-14-
Optimization of PAE allows to reach the relialbility constraints for space applications
C²S² Electrothermal modeling of HBTsTRANSISTOR G A 100TRANSISTOR GaAs
DCu
DCu
AuSnTAB
PACKAGE
100 µm25 µm280 µm
25 µm
1500 µm
1( , , )B BE BCI f V V T=⎧⎪⎨
DCuBACKSIDE
+55°C
µ
125 µmAbelfilm5025E
Package
1 .2
1 .4
1 .6
[S] ( ) ( )CE C BE BP V I V I
T j Z j Pω ω
⎧ = ⋅ + ⋅⎪⎨
= ⋅⎪⎩
2 ( , , )C BE BCI f V V T⎨ =⎪⎩
0 .0
0 .2
0 .4
0 .6
0 .8
1 .0
Ic (
A )
Réseau I(V)S
(2,2
)
[S] ( ) ( )THT j Z j Pω ω= ⋅⎪⎩
Full electrothermal model2 4 6 8 1 0 1 2 1 4 1 6 1 80 2 0
-0 .2
Vc e ( V )
freq (1.000GHz to 10.00GHz)
35
40
35
40
60
80
P)m) PAE Pout Gain
Full electrothermal model requires the knowledge of
20
25
30
15
20
25
30
15
20
40
0
PA
E (%
)
Ga
in (
dB
)
Po
ut (
dB
m PAE, Pout, Gain
( )THZ ω
-15-
2 4 6 8 10 12 14 16 180 20
Pin (dBm)
C²S²
E i t l t SIMULATION
Thermal modeling of HBTs & Packaging
32
34
Experimental set up SIMULATIONThermal Model 3D finite element simulationElectrical measurement of
the thermal impedance
20
22
24
26
28
30
1
201,00E-07 1,00E-06 1,00E-05 1,00E-04 1,00E-03 1,00E-02 1,00E-01
31
32
33
34
25
26
27
28
29
30
0 50 100 150 200
HP4194A
RT
R
RRRRR
XET
uFXKXC
.
..
.
=
+−=C=20mF
R1
R2
10k
VBE0+
DC Offset=VBE0
vbe~
Model Order ReductionModel Comparison
RR
VDD1
VDD2 1/λ11
1/λ1m
1
P1
A11.P1
A1m.P1
1
A11.T11
A1m.T1m
T11
T1
-16-
1/λ1m
1
A1m.P1
T1m
C²S² Zth extraction from electrical measurements
Principle of the measurement of the input impedance
)( TIfV = ~)(~ ∂∂),( TIfV BBE =
BE TffV ~~∂∂
+∂
=∂
PZT th ).( ∂=∂ ω~~~ IVIVP +=
BITBB ITIIB
~.~00∂∂
+∂
=∂
~ )..(~~
00 CLCEf IRVhP−=
∂
00 .. CCECCE IVIVP +=
BinISOin I
TZZ ~~
.~∂∂
+= ϕ)..(~ 00 CLCEfe
B
IRVhI∂
)..().(.~00 CLCEfethinISOin IRVhZZZ −+= ωϕ
-17-
BI∂ )()( 00 CLCEfethinISOin ϕ
C²S²Extraction of φ
Vbe (V)Ib (mA) Vbe (V)Ib (mA)
Vbe (V) T (°C)Vbe (V) T ( C)φ (V/°C)
( ) BEB
VIT
ϕ ∂=
∂BoIT∂
-18-
Ib (mA)
C²S²Input impedance variation
40
50
Re{Zin}
20
30
0
10 RL< Vce0/Ic0RL= Vce0/Ic0RL > Vce0/Ic0
I {Zi }
1E1 1E2 1E3 1E41 1E5-10
Frequency (Hz)
Im{Zin}
Frequency (Hz)
)..().(.~00 CLCEfethinISOin IRVhZZZ −+= ωϕ )..().(. 00 CLCEfethinISOin IRVhZZZ + ωϕ
Zin purely real if the condition (VCE0 - RL.IC0) =0 is verified
-19-
p y ( CE0 L C0)
C²S²
3D FE Simulation Static & Dynamic
32
34
Transient regime at the selected points
26
28
30
22
24
26
201,00E-07 1,00E-06 1,00E-05 1,00E-04 1,00E-03 1,00E-02 1,00E-01
Hot points, temperature profile
Homogeneous repartition of the power
Useful tool for the design of microwave power transistorsBut huge calculation time and heavy computing ressources
-20- 19/12/2008
But huge calculation time and heavy computing ressources
C²S² Model Order Reduction
FEMRepresentation FEMRepresentation
Linear conductivity
Real system3D Modeling
y
Numerical system
3D Modeling
Thermal subcircuit
Ci it
FTKTC =+ ..
.1/λ11A11.P1 A11.T11
Reduced Order Model MOR
Circuit synthesis
1/λ1m
1
1
P1
A1m.P1
1
A1m.T1m
T11
T1m
T1
-21-
Order Model MOR1 T1m
C²S² Comparison ANSYS vs MOR (static)
The Ritz method for MOR guarantees that steady state temperature is
-22- 19/12/2008
The Ritz method for MOR guarantees that steady state temperature is reached
C²S²
Transient regime
Comparison ANSYS vs MOR (dynamic)Transient regime
Th b f Rit t d t i th i i f th t i t iThe number of Ritz vectors determine the precision of the transient regime
Computing time (30 000 nodes) ANSYS : 10 minutes (30 V) Ritz : immediate
MOR provides a significant gain of computing time without loss of precision and allows the integration of the model in CAD softwares
-23-
precision and allows the integration of the model in CAD softwares
C²S² Influence of the mounting of the transistor Zin Ritz MORZin Ritz MOR
50Re{Zth Kovar}
30
40
{ }
50 %
20
30
Re{Zth Cu}Time constants of the transistor
0
10Im{Zth Cu}
Mounting impact
-10
0
1 10 100 1000 10000 100000 1000000
Im{Zth Kovar}
-24-
1 10 100 1000 10000 100000 1000000
C²S²
10
Low Frequency S-parameters for a HBT
0
10S
11 (
dB)
-50
0
12 (
dB)
11( )dB S
-20
-10
Mag
nitu
de S
-150
-100
Mag
nitu
de S
12( )dB S
1E2 1E3 1E4 1E5 1E6 1E71E1 1E8
-30
freq Hz
M
1E2 1E3 1E4 1E5 1E6 1E71E1 1E8
-200
freq, Hz
M
MeasuresET model
Isothermal modelfreq, Hz
0
50
1
100
200
(°)
q,
12( )Sϕ
1 0
-100
-50
Pha
se S
11
-100
0
Pha
se S
12
12( )Sϕ12( )Sϕ
1E2 1E3 1E4 1E5 1E6 1E71E1 1E8
-150
-2001E2 1E3 1E4 1E5 1E6 1E71E1 1E8
-200
freq Hz
-25- 19/12/2008
freq, Hzfreq, Hz
C²S²
Trapping EffectsTrapping Effects
-26-
C²S²GaN HEMTs Characterization
S i f G N HEMTSome issues of GaN HEMTs
- Various electrical effects (traps, thermal) which cover a large frequency band from BF to RF
- Serously impact the power behavior
Useful characterization tools:
- Pulsed I-V and S-parameters
- Load Pull frequency domain measurements
- Load Pull time domain measurements (LSNA)
-27-
C²S²
Trapping Effects (1/3)Origin: Chemical defects which induce electrical defects.
Impact: Slow current transient
Ids= f (Vgs, Vds)
Ids= f(Vgs, Vds, trapping state, t )
-28-
C²S²
Trapping Effects (2/3)
Origin: Chemical defects which induce electrical defects.
Impact: Slow current transient
Ids= f (Vgs, Vds)
Ids= f(Vgs, Vds, trapping state, t )
f t t
-29-
fast capture
C²S²
Trapping Effects (3/3)
Origin: Chemical defects which induce electrical defects.
Impact: Slow current transient
Ids= f (Vgs, Vds)
Ids= f(Vgs, Vds, trapping state, t )
F t t ( )
-30-
Fast capture (~ ns)
Slow emission (up to second)
C²S²Evidence of trapping effects
g µ
Vgs=+1.000 VVgs= +0.00nV
Gate-lag: decrease of drain current
Vgs=-8V to +1V, Vgs0=-8V, Vds0=20VVgs=-8V to +1V Vgs0=-8V Vds0=25V
Drain-lag: increase of Vknee
.8 .8Vgs=-1.000 VVgs=-2.000 VVgs=-3.000 VVgs=-4.000 VVgs=-5.000 VVgs=-6.000 VVgs=-7.000 VVgs=-8.000 VVgs=+1.000 VVgs= +0.00nVVgs=-1.000 V
Vgs=-8V to +1V, Vgs0=-8V, Vds0=25VVgs=-8V to +1V, Vgs0=-8V, Vds0=30VVgs=-8V to +1V, Vgs0=-8V, Vds0=35V
.6 .6
Vgs=-2.000 VVgs=-3.000 VVgs=-4.000 VVgs=-5.000 VVgs=-6.000 VVgs=-7.000 VVgs=-8.000 V
.4 .4
.2
0
.2
0
Gate-lag Vert (Vgs0=0V,Vds0=0V)
Drain-lagVgs0=-8 V Vds0 =15 20 25 30V
0 15 20 25 30 500 15 20 25 30 50
rouge (Vgs0=-8 V, Vds0=0V) Vgs0 8 V, Vds0 15, 20, 25, 30V
Mise en évidence des pièges τcapture << t IMPULSION << τémissionD i th l t t k l i i f d
-31-
During the pulses capture takes place, emission freezed
C²S²Nonlinear electrothermal model with trapping effects
I d(T°)
Ibk
I d(T°)
Ibk
LdRd(T°)RgdCgd
Igd(T°)RgLg LdRd(T°)RgdCgd
Igd(T°)RgLg
Vgs Vds CpdCpg Vgs Vds CpdCpg
Igs(T°) Vgs_
intGate- & Drain-lagCgs
Cds
Ids (Vgs_int(t-τ), Vds(t), T°) Igs(T°) Vg
s_intGate- & Drain-lagCgs
Cds
Ids (Vgs_int(t-τ), Vds(t), T°)
Ri
Cds
Ri
Cds
Nonlinear capacitancesRs(T°)
Ls
Transistor intrinsèque Rs(T°)
Ls
Transistor intrinsèque -Nonlinear capacitances
-Current sources
-Thermal dependence
-32-
LsLs p
-Trapping sub circuits
C²S²
Steps of the modeling process
0
5
10
15
20
dB
(S
(2,1
))
50
100
150
se (
S(2
,1))
0 7
0.80.6
0.78x75 POLAR Vgs=-4.401 V, Vds=+22.97 V, Id =+0.163 A
Vgs=+1.000 VVgs= +0.00nVVgs=-1.000 VVgs=-2.000 VVgs=-3.000 VVgs= 4 000 V 5.80E-02
5.90E-02
6.00E-02
Modèle petit-signal Modèle I-V Capacités NL Modèle thermique Modèles de pièges
Étapes de modélisation
0
5
10
15
20
dB
(S
(2,1
))
50
100
150
se (
S(2
,1))
0 7
0.80.6
0.78x75 POLAR Vgs=-4.401 V, Vds=+22.97 V, Id =+0.163 A
Vgs=+1.000 VVgs= +0.00nVVgs=-1.000 VVgs=-2.000 VVgs=-3.000 VVgs= 4 000 V
0.6
0.78x75 POLAR Vgs=-4.401 V, Vds=+22.97 V, Id =+0.163 A
Vgs=+1.000 VVgs= +0.00nVVgs=-1.000 VVgs=-2.000 VVgs=-3.000 VVgs= 4 000 V 5.80E-02
5.90E-02
6.00E-02
5.80E-02
5.90E-02
6.00E-02
Modèle petit-signal Modèle I-V Capacités NL Modèle thermique Modèles de pièges
Étapes de modélisation
,2)
5 10 15 20 25 300 35
-60
-40
-20
0
20
40
-80
60
freq, GHz
Ph
ase
(S
(1,2
))
5 10 15 20 25 300 35
-5
0
-10
freq, GHz
d
5 10 15 20 25 300 35
-24
-22
-20
-26
-18
freq, GHz
dB
(S
(1,2
))
5 10 15 20 25 300 35
0
-50
freq, GHzP
ha
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0 0
Ids
(A)
-0.1
0
0.1
0.2
0.3
0.4
0.5
-5 0 5 10 15 20 25 30 35 40 45 50
Id e
n Am
pere
s
Vds en Volts
Vgs=-4.000 VVgs=-5.000 VVgs=-6.000 VVgs=-7.000 VVgs=-8.000 V
Idss
1,55
1,6
Is_gs
2 50E 14
3,00E-14
3,50E-14
y= 1,6E-16+1,04973E-16*EXP(T/26,3157)
Cgs_1D
4,00E-13
5,00E-13
6,00E-13MesureModele
Cgd_1D
1,00E-13
1,20E-13
1,40E-13MesureModele
Vgs intRfill
k+
+
kVds
Vgs intRfill
k+
+
kVds
5.40E-02
5.50E-02
5.60E-02
5.70E-02
5.80E 02
0.00E+00 2.00E-06 4.00E-06 6.00E-06 8.00E-06 1.00E-0
,2)
5 10 15 20 25 300 35
-60
-40
-20
0
20
40
-80
60
freq, GHz
Ph
ase
(S
(1,2
))
5 10 15 20 25 300 35
-5
0
-10
freq, GHz
d
5 10 15 20 25 300 35
-24
-22
-20
-26
-18
freq, GHz
dB
(S
(1,2
))
5 10 15 20 25 300 35
0
-50
freq, GHzP
ha
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0 0
Ids
(A)
-0.1
0
0.1
0.2
0.3
0.4
0.5
-5 0 5 10 15 20 25 30 35 40 45 50
Id e
n Am
pere
s
Vds en Volts
Vgs=-4.000 VVgs=-5.000 VVgs=-6.000 VVgs=-7.000 VVgs=-8.000 V
-0.1
0
0.1
0.2
0.3
0.4
0.5
-5 0 5 10 15 20 25 30 35 40 45 50
Id e
n Am
pere
s
Vds en Volts
Vgs=-4.000 VVgs=-5.000 VVgs=-6.000 VVgs=-7.000 VVgs=-8.000 V
Idss
1,55
1,6
Is_gs
2 50E 14
3,00E-14
3,50E-14
y= 1,6E-16+1,04973E-16*EXP(T/26,3157)
Cgs_1D
4,00E-13
5,00E-13
6,00E-13MesureModele
Cgd_1D
1,00E-13
1,20E-13
1,40E-13MesureModele
Vgs intRfill
k+
+
kVds
Vgs intRfill
k+
+
kVds
5.40E-02
5.50E-02
5.60E-02
5.70E-02
5.80E 02
0.00E+00 2.00E-06 4.00E-06 6.00E-06 8.00E-06 1.00E-0
Vgs intRfill
k+
+
kVds
Vgs intRfill
k+
+
kVds
5.40E-02
5.50E-02
5.60E-02
5.70E-02
5.80E 02
0.00E+00 2.00E-06 4.00E-06 6.00E-06 8.00E-06 1.00E-0
freq (2.000GHz to 40.00GHz)
S(1
,1)
; S(2
10 20 300 400.0
Vds (V)
y = -0,0024x + 1,6085
1,2
1,25
1,3
1,35
1,4
1,45
1,5
0 20 40 60 80 100 120 140 160
0,00E+00
5,00E-15
1,00E-14
1,50E-14
2,00E-14
2,50E-14
0 20 40 60 80 100 120 140 160
y 1,6E 16 1,04973E 16 EXP(T/26,3157)
0,00E+00
1,00E-13
2,00E-13
3,00E-13
-10 -8 -6 -4 -2 0 2Vgs
Cgs
(F)
0,00E+00
2,00E-14
4,00E-14
6,00E-14
8,00E-14
-60 -50 -40 -30 -20 -10 0Vgd
Cgd
(F)
VdsVgs_int
VgsRfill
Rempty C
+
Vds
VdsVgs_int
VgsRfill
Rempty C
+
Vds
freq (2.000GHz to 40.00GHz)
S(1
,1)
; S(2
10 20 300 400.0
Vds (V)
y = -0,0024x + 1,6085
1,2
1,25
1,3
1,35
1,4
1,45
1,5
0 20 40 60 80 100 120 140 160
0,00E+00
5,00E-15
1,00E-14
1,50E-14
2,00E-14
2,50E-14
0 20 40 60 80 100 120 140 160
y 1,6E 16 1,04973E 16 EXP(T/26,3157)
0,00E+00
1,00E-13
2,00E-13
3,00E-13
-10 -8 -6 -4 -2 0 2Vgs
Cgs
(F)
0,00E+00
2,00E-14
4,00E-14
6,00E-14
8,00E-14
-60 -50 -40 -30 -20 -10 0Vgd
Cgd
(F)
VdsVgs_int
VgsRfill
Rempty C
+
Vds
VdsVgs_int
VgsRfill
Rempty C
+
Vds
VdsVgs_int
VgsRfill
Rempty C
+
Vds
VdsVgs_int
VgsRfill
Rempty C
+
Vds
Dgs=f(Vgs)
Dgd=f(Vgd)Rg
RiCdsτ
Dgs=f(Vgs)
Dgd=f(Vgd)
Dgs=f(Vgs)
Dgd=f(Vgd)Rg
RiCdsτ
Dgs=f(Vgs)
Dgd=f(Vgd)
Cgs=f(Vgs)
Cgd=f(Vgd)
Ids=f(Vgs_pièges,Vds,T)
g ( g )
Rs=f(T)
Rd f(T)
gLgCpgLsCpdLd
τGmGdCgsCgdRgd
Ids=f(Vgs,Vds,T)
Dgd f(Vgd)Ids=f(Vgs,Vds)
Cgs=f(Vgs)
Cgd=f(Vgd)
Ids=f(Vgs_pièges,Vds,T)
g ( g )
Rs=f(T)
Rd f(T)
gLgCpgLsCpdLd
τGmGdCgsCgdRgd
Ids=f(Vgs,Vds,T)
Dgd f(Vgd)Ids=f(Vgs,Vds)
Rd=f(T)
Rgd=f(T)RsRd
Rgd Rd=f(T)
Rgd=f(T)RsRd
Rgd
Various parasitics effects are successively added
-33-
Various parasitics effects are successively added
C²S² Thermal effects modelled through 3D FE simulation
HEMT 1t8x75_35µm - 7W/mm_30°CGaN1.2µm/SiC440µm/AuSn45µm/Al2mm
2224
Thermal simulations (3-5 lab) Heating as a function of time
10121416182022
Zth
[°C/
W]
2468
1.E-03 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06temps [µs]
ANSYS
FIT RC
Mise en équation avec des formes exponentielles
TEMP = 22,8.(1-e-t/τ1)+ 21 7 (1 e t/τ2)
R1 R2 R3 R4 R5
+ 21,7.(1-e-t/τ2) + 7.(1-e-t/τ3)
+ …Thermal subcircuit
C1 C2 C3 C4 C5
-34-
I = Pdissipée U=Tchuck_°C
T °C
C²S²Transistor parameters dependence on temperature
I-V @ various temperatures
Thermal laws
0 0049 0 68891.4
1.6
Rd0.5
0.6
0.78x75 POLAR Vgs=+0.163 V, Vds=+10.97mV, Id =-0.550mA
Vgs=+1.000 VVgs= +0.00nVVgs=-1.000 VVgs=-2.000 VVgs=-3.000 VVgs=-4.000 VVgs=-5.000 VVgs=-6.000 V
p
Equations
y = 0.0029x + 0.6375
y = 0.0049x + 0.6889
0.4
0.6
0.8
1
1.2
Rs,
Rd
Rs
0
0.1
0.2
0.3
0.4
Id e
n A
mpe
res
gVgs=-7.000 VVgs=-8.000 V
Thermal parameters- Access resistances
0 50 100 150 200T°C
-0.1-5 0 5 10 15 20 25 30 35 40 45
Vds en Volts
IV @ 25°C... Rs= Rs0+ α _Rs.TRd= Rd + α Rd T
...- Current sources- Diodes
y = -0.0008x + 1.1543
1 061.081.1
1.121.141.161.18
Idss
0 3
0.4
0.5
0.6
0.7
mpe
res
8x75 POLAR Vgs=+0.165 V, Vds=+9.709mV, Id =-0.440mA
Vgs=+1.000 VVgs= +0.00nVVgs=-1.000 VVgs=-2.000 VVgs=-3.000 VVgs=-4.000 VVgs=-5.000 VVgs=-6.000 VVgs=-7.000 VVgs=-8.000 V
Rd= Rd0+ α_Rd.TIdss= Idss0+ Idsst.TP=P0+Pt.TNgs=Ngs0+Ngst.T
1.021.041.06
0 50 100 150 200T°C
-0.1
0
0.1
0.2
0.3
-5 0 5 10 15 20 25 30 35 40 45
Id e
n A
m
Vds en Volts
Ngd=Ngd0+Ngdt.TIsgs=Isgs0+Isgst.e (T/Tsgs)
Isgd=Isgd0+Isgdt.e (T/Tsgd).
-35-
Vds en Volts
IV @ 150°C..
C²S² Topology of the trapping effects model
ff f ( )Trapping effects modify the gate command (back gating)
transients on Vgs Transients of the drain current
Ch f th it I i d tCharge of the capacitance= Ionized traps
Charge through Rcapture, Emission through Rémission
Tuning of the magnitude of theDiode= dissymetry of the capture and emission process
Tuning of the magnitude of the trapping effects
Fundamental assumption : dissymetry of the capture and emission process
-36-
Fundamental assumption : dissymetry of the capture and emission process
C²S²The model takes the knee walk out into account
1.2Measured Drain LagSimulations drain-lag ON/OFF
0.8
1.0
mm
)
0 6
0.8
1.0
1.2
mm
)
Vgs0=-7V, Vds0=25VVgs0=-7V, Vds0=0V
0.4
0.6
Ids
(A/m
0 0
0.2
0.4
0.6
Ids
(A/m
0.0
0.2
0 10 20 30 40 50Vds (V)
0.00 10 20 30 40 50Vds (V)
Pulsed Measurements
V 0 7V Vd 0 0V (bl )- Vgs0=-7V, Vds0=0V (bleu)
- Vgs0=-7V, Vds0=25V (rouge) Simulated I-V Characteristics
@ Vgs0=-7V, Vds0=0V (pointillés)
@ V 0 7V Vd 0 25V ( )@ Vgs0=-7V, Vds0=25V (rouge)
Measure @ Vgs0=-7V, Vds0=25V (bleu)
-37-
C²S²
Load pull measurements at various loads
Large signal impact of traps (1)Load pull measurements at various loads
1 : ZOPT2 Z1 2
34 1 : ZOPT
2 Z1 2
34
2 : Z2_VSWR=2.53 : Z3_VSWR=2.54 : Z4 VSWR 1 6
1 25 6
2 : Z2_VSWR=2.53 : Z3_VSWR=2.54 : Z4 VSWR 1 6
1 25 6
4 : Z4_VSWR=1.65 : Z5_VSWR=2.56 : Z6 VSWR=2.5
4 : Z4_VSWR=1.65 : Z5_VSWR=2.56 : Z6 VSWR=2.5__
Class AB, Vds=25 V , DC Bias , RF CW, 10 GHz
-38-
C²S²
L d t d f ti
Large signal impact of traps (2)
Load
Load tuned for optimum power
0.8
_int
.i
Zlo
ad
0.2
0.4
0.6
0 0
cycl
ere
seau
_IV
_sdd
..Ids
_Pièges ONPièges OFFMesure
(0.000 to 0.000)
-160
-155
a_
in)
40 250
10 20 30 40 50 600 70
0.0
indep(cycle)X1.Vd_int
-170
-165
Ph
ase
(G
am
ma
20
0
PA
E(%
)
200
150ID
S (
mA
)
0.95
in)
-5 0 5 10 15 20-10 25
-175
Pin (dBm)
-5 0 5 10 15 20-10 250
Pin (dBm)
-5 0 5 10 15 20-10 25150
Pin (dBm)316
0.85
0.90
Ma
g (
Ga
mm
a_
i
1
2
Po
ut (
W)
12
14
10
Ga
in(d
B)
-39-
-5 0 5 10 15 20-10 25
0.80
Pin (dBm)
0.0
2
0.0
4
0.0
6
0.0
8
0.1
0
0.1
2
0.1
4
0.1
6
0.1
8
0.0
0
0.2
0
0
Pin (W)
-5 0 5 10 15 20-10 2510
Pin (dBm)
C²S²Explanantion of the decrease of the average current
220
180
200
DS
(m
A)
0 -10 -5 0 5 10 15-15 20
180
160
ID
Pin (dBm)2 5
-40-
C²S²Explanantion of the decrease of the average current
220
180
200
DS
(m
A)
0 -10 -5 0 5 10 15-15 20
180
160
ID
Pin (dBm)2 5
-41-
C²S²Explanantion of the decrease of the average current
220
180
200
DS
(m
A)
0 -10 -5 0 5 10 15-15 20
180
160
ID
Pin (dBm)2 5
-42-
C²S²Explanantion of the decrease of the average current
220
180
200
DS
(m
A)
0 -10 -5 0 5 10 15-15 20
180
160
ID
Pin (dBm)2 5
-43-
C²S²Validation of the model with mismatched loads
Load
TOS=1,6
0.6
0.8
s_in
t.i
Zlo
ad
0.2
0.4
0.6
0 0
cycl
ere
seau
_IV
_sdd
..IdPièges ON
Pièges OFFMesure
-160
_in
)
40 220
(0.000 to 0.000)
10 20 30 40 50 600 70
0.0
indep(cycle)X1.Vd_int
-170
-165
Ph
ase
(G
am
ma
_
20
PA
E(%
)
180
200
IDS
(m
A)
0.98
)
-10 -5 0 5 10 15-15 20
-175
Pin (dBm)
-10 -5 0 5 10 15-15 200
Pin (dBm)
-10 -5 0 5 10 15-15 20160
Pin (dBm)
2 0
2.522
0.92
0.94
0.96
Ma
g (
Ga
mm
a_
in)
1.0
2.0
Po
ut (
W)19
Ga
in(d
B)
-44-
-10 -5 0 5 10 15-15 20
0.90
Pin (dBm)
M
0.0
1
0.0
2
0.0
3
0.0
4
0.0
5
0.0
6
0.0
7
0.0
8
0.0
0
0.0
9
0.0
Pin (W)
-10 -5 0 5 10 15-15 2014
Pin (dBm)
C²S²Validation of the model with mismatched loads
Load
TOS=2,50.6
0.8
ed..Id
s_in
t.i
Zlo
ad
10 20 30 40 50 600 70
0.2
0.4
0.0
cycl
ere
seau
_IV
_sddPièges ON
Pièges OFFMesure
-155
-150
a_
in)
20
30
200
220
)
(0.000 to 0.000)indep(cycle)
X1.Vd_int
-170
-165
-160
Ph
ase
(G
am
ma
10
20
0
PA
E(%
)
180
200
160
IDS
(m
A)
0.85
n)
-5 0 5 10 15 20 25-10 30
-175
Pin (dBm)
-5 0 5 10 15 20 25-10 300
Pin (dBm)
-5 0 5 10 15 20 25-10 30160
Pin (dBm)
2.0
2.512
0.75
0.80
Ma
g (
Ga
mm
a_
i
1.0
2.0
Po
ut (
W)
8
10
6
Ga
in(d
B)
-45-
-5 0 5 10 15 20 25-10 30
0.70
Pin (dBm)
0.1 0.2 0.3 0.40.0 0.50.0
Pin (W)
-5 0 5 10 15 20 25-10 306
Pin (dBm)
C²S²
Measurements @ 5 GHz 25V dc/cw
LSNA Measurements
Measurements @ 5 GHz, 25V, dc/cw
0.4
0.5
0.1
0.2Pièges ONPièges OFFM
-0.0
0.1
0.2
0.3
Ids
-0.2
-0.1
0.0
Igs
TOS 4
5dB i
Mesure
10 20 30 40 500 60-0.1
Vds-10 -8 -6 -4 -2 0-12 2
-0.3
Vgs
0 6
0.8
0 2
0.3
5dB compression
-0.2
0.0
0.2
0.4
0.6
Ids
0 2
-0.1
0.0
0.1
0.2
Igs
TOS 3,3
320 0.8
0 10 20 30 40 50 60-10 70
0.2
-0.4
Vds-10 -8 -6 -4 -2 0-12 2
-0.2
-0.3
Vgs
7dB compression
220240260280300
Ids
(mA
)
0.2
0.4
0.6
Ids
TOS 2
-46--5 0 5 10 15 20-10 25
200220
180
Pe (dBm)
10 15 20 25 30 355 40
0.0
-0.2
Vds
TOS 2
8dB compression
C²S²Design of a 2 stages AlGaN/GaN MMIC HPA Output power PAE & Gain at 9 GHz
45
50) 35
40
P t
Output power, PAE & Gain at 9 GHz Drain Bias 32V
30
35
40
45m
) & g
ain
(dB
)
20
25
30
35
E (%
)
Pout
PAE
15
20
25
30
Pow
er (d
Bm
5
10
15
20
PAE
gain
1010 15 20 25 30 35
Input Power (dBm)
0
• Pout = 47.7 dBm ( 58 W)• PAE = 38 %
Chip size : 16.5 mm²4300x3800 µm²
1st stage : 2.4 mm• 6.5 W/mm • Gain = 14.6 dB• Vds = 32V Ids = 2 3A
State-of-the-Art Output Power with AlG N/G N MMIC HPA t X B d
g2nd stage : 8.96 mm
-47-
• Vds0 = 32V, Ids0 = 2.3A AlGaN/GaN MMIC HPA at X Band
C²S²
Impact IonizationImpact Ionization
-48-
C²S² Impact Ionization in GaAs HEMTs
0.15
0.24x75 POLAR Vgs=-1.481 V, Vds=+6.070 V, Id =+41.25µA
Vgs=+0.750 VVgs=+0.750 VVgs=+0.750 VVgs=+0.750 VVgs=+0.750 VVgs=+0.750 VVgs=+0.750 VVgs=+0.750 VVgs=+0.750 V[ ]22Yℜ
0.2A
0 05
0.1
Id e
n Am
pere
s
[ ]22
( )DS DSI V0
0.05 ( )DS DSI V
-0.050 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5
Vds en Volts 5V
Interaction between II and Traps
[ ]22YℑInteraction between II and Traps
Temperature dependence
Frequency dispersion of output[ ]
Frequency dispersion of output
conductance
-49- 19/12/2008
C²S²
I t i i ti i f d d t
PHEMT GaAs/GaInAs Model for Impact Ionization
Impact ionization is frequency dependent Model with impact ionization source
Model with impact ionization filtered around 2 GHz
-50-
Lowpass filter with a cut-off frequency of 2 GHz
C²S² Conclusions
Dispersive effects have a strong impact on transistors performances
Th l Eff i ll d iThermal Effects in all devicesTrapping effects in HEMTs (GaN and GaAs)Impact ionization effects coupled with trapping effectsImpact ionization effects coupled with trapping effects
Require specialized characterization tools Pulsed I-V and S-parameters measurementspLoad Pull Frequency and Time DomainsLow Frequency Characterization Ph i l d th l i l tiPhysical and thermal simulation
Need further investigations for checking the consistency of different kinds of characterization and modelingdifferent kinds of characterization and modeling.
To provide usefull tools for the technology assessment and improvement as well as optimization of PA performances
-51- 19/12/2008
p p p