Download Effect of Source Bonding Wires in HEMT devices: Probe Station

Effect of Source Bonding Wires in
HEMT devices: Probe Station
Measurements
J.D. Gallego, C. Diez González, I. López, I. Malo
IT-CDT 2016-18
Observatorio de Yebes
Apdo. 148 19080 Guadalajara
SPAIN
Phone: +34 949 29 03 11
Fax: +34 949 29 00 63
YebeS
Observatorio de Yebes
Apartado 148, 19080 Guadalajara, SPAIN
Effect of Source Bonding Wires in HEMT devices:
Probe Station Measurements
Change Record
Revision
Date
Affected Paragraphs(s)
A
2016-11-02
All
Rev. A
Reason/Initiation/Remarks
First Issue
IT-CDT 2016-18
Page 2 of 22
YebeS
Observatorio de Yebes
Apartado 148, 19080 Guadalajara, SPAIN
Effect of Source Bonding Wires in HEMT devices:
Probe Station Measurements
TABLE OF CONTENTS
1.
2.
3.
4.
5.
6.
7.
Introduction ......................................................................................................................... 4
Measurement equipment ..................................................................................................... 4
Calibration ........................................................................................................................... 4
Measurements ...................................................................................................................... 5
Conclusions ......................................................................................................................... 6
Appendix I ........................................................................................................................... 9
Appendix II........................................................................................................................ 10
Rev. A
IT-CDT 2016-18
Page 3 of 22
YebeS
Observatorio de Yebes
Apartado 148, 19080 Guadalajara, SPAIN
Effect of Source Bonding Wires in HEMT devices:
Probe Station Measurements
1. Introduction
This report presents the results of the measurement of several types of HEMTs including the
effect of bonding wires. The objective was to asset the accuracy of the models used for the
design of cryogenic amplifiers. The measurements were carried out in a Probe Station in the
0.250-110 GHz frequency range at ambient temperature using the new equipment available in
Yebes. The most interesting finding of this work is the resonances which appear at high
frequencies (typically >50 GHz) and that were not predicted by the simple models based on
equivalent circuits obtained by fitting on-wafer measurements of S parameters of the
transistors plus ideal inductors to simulate bonding wires to the chip. These resonances can be
explained by the parasitic capacitance of the source metallization and bonding pads of the
transistors which is not usually included in the equivalent circuit. This effect appears at high
frequency and does not affect much to the S parameters below 20 GHz, but can be dramatic at
higher frequency, since it can produce oscillations, making the transistors virtually unusable
in some cases.
2. Equipment









Probe station mod. MPS 150 (Cascade Microtech)
Coplanar probes mod. ACP 110-A-GSG-125 (Cascade Microtech)
Vector network analyzer mod. PNA-X 5247 (Keysight)
Millimeter wave controller mod. N5261A (Keysight)
Millimeter wave heads mod. N5250CX10 (Keysight)
Power supply mod. N3280A (Agilent)
Transitions from coplanar to microstrip mod. ProbePoint 0503 (Jmicro)
Coplanar calibration substrate (ISS) mod. 104-783A (Cascade Microtech)
Microstrip calibration substrate mod. CM05LX (Jmicro)
3. Calibration
The chip measurements were taken with a standard LRRM calibration with the Cascade
Microtech calibration substrate (Impedance Standard Substrate, figure 3) using WinCal
software. With the standards used this calibration performs reasonably well in all the
250MHz-110 GHz range used. It was verified with an open (probes in air) and with a long
matched coplanar line (~27ps) in the ISS substrate. The ISS was used in combination with an
absorbing ISS holder (SN 116-334) as recommended by Cascade.
A microstrip calibration with the Jmicro calibration substrate (figure 2) and the multiline TRL
(NIST type) algorithm implemented in WinCal was attempted, but it was not possible to
calibrate over 90 GHz due to the limitation of the minimum line length in the Jmicro
substrate. However, the microstrip lines over 5 mil alumina still perform reasonably well up
to 110 GHz, as checked with measurements of the Jmicro thru (calibrated with the Cascade
ISS in the coplanar reference plane). Finally, it was decided to take the microstrip
measurements with the coplanar LRRM calibration but de-embedding the effect of the two
coplanar to microstrip transitions. This was performed with a built-in feature of the PNA-X
which allows de-embedding circuits characterized by their S-parameter files. Appendix I
contain some information of the models used to generate the files used for de-embedding.
Rev. A
IT-CDT 2016-18
Page 4 of 22
YebeS
Observatorio de Yebes
Apartado 148, 19080 Guadalajara, SPAIN
Effect of Source Bonding Wires in HEMT devices:
Probe Station Measurements
4. Measurements
Different types of devices were tested with and without bonding wires with a bias point near
the optimum expected for minimum noise at ambient temperature. The details of the results of
the measurements and of the parameters of the models of the equivalent circuits are shown in
Table I and Appendix II. The measurements without bonding wires were taken with the chips
on the Gel-Pack box used for storage. It was attempted to measure the chips mounted in the
brass plate with the coplanar to microstrip transitions in place but without bonding wires.
However, that measurement was not good since there was a strong coupling of the coplanar
probes and the adjacent microstrip lines (transitions) terminated in open circuit. This appeared
in the S parameters as a resonance which spoiled the measurement.
In the case of ETH 150 end ETH 50 devices, the chip measurements were compared with the
equivalent circuit provided by ETH which includes pad parasitic elements obtained by
measurement of dummy structures. The new measurements agree very well with ETH model
predictions. Note that the maximum frequency used initially by ETH to fit the model
parameters was 40 GHz and the new measurements are taken up to 110 GHz. Obviously a
better result could be obtained by tuning some equivalent circuit parameters to fit the new
data but it was preferred to leave the original values to illustrate the small difference. The chip
equivalent circuit parameters are kept unaltered for the comparison with the measurements
with bonding wires. Only the values of Lg, Ld and Ls are re-optimized to fit the measured
data. In addition, a source pad capacitor (CpadS in figure 1) is introduced in an attempt to
simulate the resonances found in the measurement.
Not such an elaborated chip equivalent circuit was available for the HRL 150 or the IAF 150.
The HRL chip could be measured on the Gel Pack and the values of the equivalent circuit
parameters were obtained by fitting to the measurements in the 0.25-110 GHz range. The
IAF 150 is a smaller chip with a pitch of 100 um, and could not be directly probed in the
present setup (configured for 125 um pitch). The equivalent circuit parameters of the IAF 150
device were obtained by fitting to the data obtained with three bonding wires on each side.
Figure 1: Equivalent circuit used for the transistors including bonding pad parasitics. Note
the presence of CpadS (Source bonding pad) which was introduced to model the
resonances observed in the measurements with bonding wires.
Rev. A
IT-CDT 2016-18
Page 5 of 22
YebeS
Observatorio de Yebes
Apartado 148, 19080 Guadalajara, SPAIN
Effect of Source Bonding Wires in HEMT devices:
Probe Station Measurements
5. Conclusions
In the results presented in Appendix II is clearly visible the problem which appears due the
resonance of the parasitic capacitance of the source pads with the bonding wires. Table I
shows the values of the parasitic capacitance extracted from the measurements. The fitting of
the models to the measured S parameters is far from perfect, especially above the resonance.
Nevertheless, the qualitative behavior is more or less predicted by this simple model.
Attempts to improve the fitting with more complex topologies did not succeed in obtaining a
better result.
The relative values of the source pad parasitic capacitance obtained (Table I) are in agreement
with what one would expect from the geometry of the transistors and pads. For example,
ETH 150-1 and ETH 150-2 are identical in everything but the height of the chip and the
capacitance scales (inversely) in almost the same factor as the height (~2). The highest value
of parasitic capacitance is obtained for the IAF 150 HEMT which is the one with the smaller
chip height. Looking at the photos of the devices it seems clear that the larger area of source
pads corresponds to the ETH layouts. This suggests that the performance of these devices
(respect to resonance with bonding wires) could be improved by modifying the layout,
reducing the source metallization area as much as possible. The HRL 150 is the device in
which the resonance effect appears at higher frequency. Coincidentally it is also the device
with a smaller layout. This could explain the remarkable stable performance of the HRL 150
in cryogenic amplifiers, although it is true that the low value of transconductance also helps in
this.
The ETH 50 is particularly problematic since it becomes clearly unstable (|S11| >1) with
bonding wires for frequencies above the resonance. It will be difficult to use this device in a
cryogenic amplifier unless an extremely low source inductance of the bonding wires can be
obtained in the final configuration. Note that in present measurements, even with three short
bonding wires on each side, the device is marginally unstable. More tests are needed to check
whether an inductance low enough can be achieved to avoid this effect in a practical
configuration.
Rev. A
IT-CDT 2016-18
Page 6 of 22
YebeS
Observatorio de Yebes
Apartado 148, 19080 Guadalajara, SPAIN
Effect of Source Bonding Wires in HEMT devices:
Probe Station Measurements
TABLE I
Gate Width (um)
Chip Height (um)
REF
Vd (V)
Id (mA)
ETH
150-1
150
~95
T-245
0.5
15
ETH
150-2
150
~170
T-273
0.5
15
ETH
50
50
~120
T-317
0.5
5
HRL
150
150
~120
T-78
0.75
15
IAF
150
150
50
T-315
0.5
15
PAD PARASITICS
Cpg1 (pF)
Cpg2 (pF)
Rgsub1 (Ω)
Rgsub2 (Ω)
Rgsub3 (Ω)
Cpd1 (pF)
Cpd2 (pF)
Rdsub1 (Ω)
Rdsub2 (Ω)
Rdsub3 (Ω)
0.0162
0.0006
14.1
42K
392K
0.0176
0.0009
13.8
74K
318K
0.0162
0.0006
14.1
42K
392K
0.0176
0.0009
13.8
74K
318K
0.0146
0.0070
10.3
98K
90K
0.0155
0.0013
9.6
73K
87K
-
-
CHIP Eq. CKT
Rg (Ω)
Rd (Ω)
Rs (Ω)
Cgs (pF)
Rgs (Ω)
Cds (pF)
Rds (Ω)
Cgd (pF)
Gm (mS)
Tau (ps)
1.9
1.5
1.2
0.0984
3.8
0.0397
59.1
0.0288
187.5
0.064
1.9
1.5
1.2
0.0984
3.8
0.0397
59.1
0.0288
187.5
0.064
0.6
4.7
4.7
0.0276
7.9
0.0155
148.2
0.0112
62.1
0
0.3
1.4
0.5
0.1075
2.9
0.0471
83.62
0.0321
129.9
0.079
0.3
1.4
0.5
0.0805
3.1
0.0467
50.41
0.0431
162.3
0
CHIP
Lg (nH)
Ld (nH)
Ls (nH)
CpadS (pF)
0.0653
0.0528
0.0013
-
0.0653
0.0528
0.0013
-
0.046
0.050
0.0006
-
0.0218
0.0250
0.0051
-
-
3 BW
Lg (nH)
Ld (nH)
Ls (nH)
CpadS (pF)
0.0993
0.1215
0.0413
0.1470
0.1294
0.1614
0.0471
0.0796
0.1159
0.1012
0.0513
0.1196
0.1428
0.1258
0.0376
0.1234
0.1054
0.1151
0.0256
0.2020
1 BW
Parameters of the equivalent circuit of the transistors measured
Lg (nH)
Ld (nH)
Ls (nH)
CpadS (pF)
0.1114
0.1107
0.0792
0.1524
0.1297
0.1469
0.1062
0.0700
0.1159
0.1012
0.1039
0.1078
0.1168
0.1083
0.0758
0.0929
0.1020
0.1073
0.0664
0.1601
Rev. A
IT-CDT 2016-18
Page 7 of 22
YebeS
Observatorio de Yebes
Apartado 148, 19080 Guadalajara, SPAIN
Effect of Source Bonding Wires in HEMT devices:
Probe Station Measurements
Figure 2: Jmicro coplanar to microstrip transition and calibration substrate. The frequency
range of the NIST multiline TRL calibration is limited by the length of the
shortest line to ~80-90 GHz.
Figure 3: Calibration substrate used for LRRM calibration on the coplanar reference plane
in the 0.250-110 GHz frequency range.
Rev. A
IT-CDT 2016-18
Page 8 of 22
Observatorio de Yebes
Apartado 148, 19080 Guadalajara, SPAIN
YebeS
Effect of Source Bonding Wires in HEMT devices:
Probe Station Measurements
6. Appendix I
Model used for de-embedding J-micro coplanar to microstrip transitions.
(Half of the model is used to generate S2P files used for de-embedding)
S2P
SNP1
File="Thru_Jmicro_cal_ISS_cascade_2.s2p"
Frequencies set f or
well behav ed time domain
1
Term
Term2
Num=2
Z=50 Ohm
Term
Term1
Num=1
Z=50 Ohm
S_Param
SP1
Start=(0.274314214/2) GHz
Stop=109.999999814 GHz
Step=0.274314214 GHz
Var
Eqn
Jmicro thru
(with problems)
measured PNA-X
C
TLINP
C1
C=Cpad pF TL2
Z=51 Ohm
L=len mm
K=1.001
A=att
F=1 GHz
TanD=0.0001
Mur=1
TanM=0
Sigma=0
S2P_Eqn
S2P1
S[1,1]=
S[1,2]=polar(sqrt(CALC),1*Pcor/2)
S[2,1]=
S[2,2]=
Z[1]=
Z[2]=
VAR
Term1
Cpad=0.004 {t}
len=1.45 {t}
att=0 {t}
S2P_Eqn
S2P2
S[1,1]=
S[1,2]=polar(sqrt(CALC),1*Pcor/2)
S[2,1]=
S[2,2]=
Z[1]=
Z[2]=
AMPLITUDE AND PHASE CORRECTION
S(4,3)
S(2,1)
S(3,3)
S(1,1)
Term
Term3
Num=3
Z=50 Ohm
2
Re f
S-PARAMETERS
-1.0
-0.8
-0.6
freq (137.2MHz to 110.0GHz)
0
-0.2
0.0
0.2
0.4
0.6
Term
Term4
Num=4
Z=50 Ohm
0.8
1.0
100
110
freq (137.2MHz to 110.0GHz)
0.0
-10
-0.4
C
C2
C=Cpad pF
TLINP
TL3
Z=51 Ohm
L=len mm
K=1.001
A=att
F=1 GHz
TanD=0.0001
Mur=1
TanM=0
Sigma=0
-0.5
-1.0
-30
dB(S(3,4))
dB(S(1,2))
dB(S(1,1))
dB(S(3,3))
-20
-40
-50
-1.5
-2.0
-60
-2.5
-70
-80
-3.0
0
10
20
30
40
50
60
70
80
90
100
110
freq, GHz
Rev. A
0
10
20
30
40
50
60
70
80
90
freq, GHz
IT-CDT 2016-18
Page 9 of 22
YebeS
Observatorio de Yebes
Apartado 148, 19080 Guadalajara, SPAIN
Effect of Source Bonding Wires in HEMT devices:
Probe Station Measurements
7. Appendix II
Comparison of measurements and models of the transistors tested.
Rev. A
IT-CDT 2016-18
Page 10 of 22
Observatorio de Yebes
Apartado 148, 19080 Guadalajara, SPAIN
YebeS
Effect of Source Bonding Wires in HEMT devices:
Probe Station Measurements
Smodel(1,2)
Smeas(1,2)
Smodel(1,1)
Smeas(1,1)
ETH 150 (chip height: 95 um)
-0.15
freq (1.000GHz to 110.0GHz)
-0.10
-0.05
0.00
0.05
0.10
0.15
freq (1.000GHz to 110.0GHz)
DDS_File_Name
-10
-8
-6
-4
-2
0
2
4
6
8
10
Smeas(2,2)
Smodel(2,2)
Smeas(2,1)
Smodel(2,1)
ETH_150_ISS_orig
freq (1.000GHz to 110.0GHz)
freq (1.000GHz to 110.0GHz)
Rev. A
IT-CDT 2016-18
Page 11 of 22
Observatorio de Yebes
Apartado 148, 19080 Guadalajara, SPAIN
YebeS
Effect of Source Bonding Wires in HEMT devices:
Probe Station Measurements
ETH 150 (chip height: 95 um) 3 BW each side
m1
freq=59.00GHz
Smodel(1,2)
Smeas(1,2)
Smodel(1,1)
Smeas(1,1)
m1
-1.5
freq (1.000GHz to 80.00GHz)
-1.0
-0.5
0.0
0.5
1.0
1.5
freq (1.000GHz to 80.00GHz)
DDS_File_Name
ETH_150_3BW_orig
-10
-8
-6
-4
-2
0
2
4
6
8
10
freq (1.000GHz to 80.00GHz)
freq (1.000GHz to 80.00GHz)
Rev. A
m2
freq=51.00GHz
Smeas(2,2)
Smodel(2,2)
Smeas(2,1)
Smodel(2,1)
m2
IT-CDT 2016-18
Page 12 of 22
Observatorio de Yebes
Apartado 148, 19080 Guadalajara, SPAIN
YebeS
Effect of Source Bonding Wires in HEMT devices:
Probe Station Measurements
ETH 150 (chip height: 95 um) 1 BW each side
Smodel(1,2)
Smeas(1,2)
Smodel(1,1)
Smeas(1,1)
m1
freq=35.00GHz
m1
-1.5
freq (1.000GHz to 80.00GHz)
-1.0
-0.5
0.0
0.5
1.0
1.5
freq (1.000GHz to 80.00GHz)
DDS_File_Name
ETH_150_1BW_orig
-10
-8
-6
-4
-2
0
2
4
6
8
10
Smeas(2,2)
Smodel(2,2)
Smeas(2,1)
Smodel(2,1)
m2
freq=21.00GHz
freq (1.000GHz to 80.00GHz)
freq (1.000GHz to 80.00GHz)
Rev. A
m2
IT-CDT 2016-18
Page 13 of 22
Observatorio de Yebes
Apartado 148, 19080 Guadalajara, SPAIN
YebeS
Effect of Source Bonding Wires in HEMT devices:
Probe Station Measurements
Smodel(1,2)
Smeas(1,2)
Smodel(1,1)
Smeas(1,1)
ETH 150 (chip height: 170 um) 3 BW each side
m1
m1
freq=75.00GHz
-1.5
freq (1.000GHz to 80.00GHz)
-1.0
-0.5
0.0
0.5
1.0
1.5
freq (1.000GHz to 80.00GHz)
DDS_File_Name
ETH_150_h160_3BW_...
-10
-8
-6
-4
-2
0
2
4
6
8
10
Smeas(2,2)
Smodel(2,2)
Smeas(2,1)
Smodel(2,1)
m2
freq (1.000GHz to 80.00GHz)
freq (1.000GHz to 80.00GHz)
Rev. A
m2
freq=75.00GHz
IT-CDT 2016-18
Page 14 of 22
Observatorio de Yebes
Apartado 148, 19080 Guadalajara, SPAIN
YebeS
Effect of Source Bonding Wires in HEMT devices:
Probe Station Measurements
m1
m1
freq=57.00GHz
Smodel(1,2)
Smeas(1,2)
Smodel(1,1)
Smeas(1,1)
ETH 150 (chip height: 170 um) 1 BW each side
-1.5
freq (1.000GHz to 80.00GHz)
-1.0
-0.5
0.0
0.5
1.0
1.5
freq (1.000GHz to 80.00GHz)
DDS_File_Name
-10
-8
-6
-4
-2
0
2
4
6
8
10
Smeas(2,2)
Smodel(2,2)
Smeas(2,1)
Smodel(2,1)
ETH_150_h160_1BW_...
m2
freq=41.00GHz
freq (1.000GHz to 80.00GHz)
freq (1.000GHz to 80.00GHz)
Rev. A
m2
IT-CDT 2016-18
Page 15 of 22
Observatorio de Yebes
Apartado 148, 19080 Guadalajara, SPAIN
YebeS
Effect of Source Bonding Wires in HEMT devices:
Probe Station Measurements
Smodel(1,2)
Smeas(1,2)
Smodel(1,1)
Smeas(1,1)
HRL 150 (chip height: 120 um)
-0.15
freq (1.000GHz to 110.0GHz)
-0.10
-0.05
0.00
0.05
0.10
0.15
freq (1.000GHz to 110.0GHz)
DDS_File_Name
-8
-6
-4
-2
0
2
4
6
8
Smeas(2,2)
Smodel(2,2)
Smeas(2,1)
Smodel(2,1)
HRL_150_ISS
freq (1.000GHz to 110.0GHz)
freq (1.000GHz to 110.0GHz)
Rev. A
IT-CDT 2016-18
Page 16 of 22
Observatorio de Yebes
Apartado 148, 19080 Guadalajara, SPAIN
YebeS
Effect of Source Bonding Wires in HEMT devices:
Probe Station Measurements
HRL 150 (chip height: 120 um) 3 BW each side
Smodel(1,2)
Smeas(1,2)
Smodel(1,1)
Smeas(1,1)
m1
m1
freq=74.00GHz
-1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8
freq (1.000GHz to 80.00GHz)
1.0 1.2
freq (1.000GHz to 80.00GHz)
DDS_File_Name
HRL_150_3BW
-8
-6
-4
-2
0
2
4
6
8
Smeas(2,2)
Smodel(2,2)
Smeas(2,1)
Smodel(2,1)
m2
m2
freq=69.00GHz
freq (1.000GHz to 80.00GHz)
freq (1.000GHz to 80.00GHz)
Rev. A
IT-CDT 2016-18
Page 17 of 22
Observatorio de Yebes
Apartado 148, 19080 Guadalajara, SPAIN
YebeS
Effect of Source Bonding Wires in HEMT devices:
Probe Station Measurements
HRL 150 (chip height: 120 um) 1 BW each side
m1
freq=64.00GHz
Smodel(1,2)
Smeas(1,2)
Smodel(1,1)
Smeas(1,1)
m1
-1.5
freq (1.000GHz to 80.00GHz)
-1.0
-0.5
0.0
0.5
1.0
1.5
freq (1.000GHz to 80.00GHz)
DDS_File_Name
HRL_150_1BW
-8
-6
-4
-2
0
2
4
6
8
Smeas(2,2)
Smodel(2,2)
Smeas(2,1)
Smodel(2,1)
m2
freq=71.00GHz
freq (1.000GHz to 80.00GHz)
freq (1.000GHz to 80.00GHz)
Rev. A
m2
IT-CDT 2016-18
Page 18 of 22
Observatorio de Yebes
Apartado 148, 19080 Guadalajara, SPAIN
YebeS
Effect of Source Bonding Wires in HEMT devices:
Probe Station Measurements
ETH 50 (chip height: 120 um) 3 BW each side
Smodel(1,2)
Smeas(1,2)
Smodel(1,1)
Smeas(1,1)
m1
freq=47.00GHz
m1
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
freq (1.000GHz to 80.00GHz)
freq (1.000GHz to 80.00GHz)
DDS_File_Name
ETH_50_Jmicro_deemb...
-4
-3
-2
-1
0
1
2
3
4
Smeas(2,2)
Smodel(2,2)
Smeas(2,1)
Smodel(2,1)
m2
freq=45.00GHz
freq (1.000GHz to 80.00GHz)
freq (1.000GHz to 80.00GHz)
Rev. A
m2
IT-CDT 2016-18
Page 19 of 22
Observatorio de Yebes
Apartado 148, 19080 Guadalajara, SPAIN
YebeS
Effect of Source Bonding Wires in HEMT devices:
Probe Station Measurements
ETH 50 (chip height: 120 um) 1 BW each side
Smodel(1,2)
Smeas(1,2)
Smodel(1,1)
Smeas(1,1)
m1
freq=39.00GHz
m1
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
freq (1.000GHz to 80.00GHz)
freq (1.000GHz to 80.00GHz)
DDS_File_Name
ETH_50_Jmicro_deemb...
-4
-3
-2
-1
0
1
2
3
4
Smeas(2,2)
Smodel(2,2)
Smeas(2,1)
Smodel(2,1)
m2
freq=27.00GHz
freq (1.000GHz to 80.00GHz)
freq (1.000GHz to 80.00GHz)
Rev. A
m2
IT-CDT 2016-18
Page 20 of 22
Observatorio de Yebes
Apartado 148, 19080 Guadalajara, SPAIN
YebeS
Effect of Source Bonding Wires in HEMT devices:
Probe Station Measurements
IAF 150 (chip height: 50 um) 3 BW each side
Smodel(1,2)
Smeas(1,2)
Smodel(1,1)
Smeas(1,1)
m1
m1
freq=71.00GHz
-1.5
freq (1.000GHz to 80.00GHz)
-1.0
-0.5
0.0
0.5
1.0
1.5
freq (1.000GHz to 80.00GHz)
DDS_File_Name
IAF_150_3BW_test
-8
-6
-4
-2
0
2
4
6
8
Smeas(2,2)
Smodel(2,2)
Smeas(2,1)
Smodel(2,1)
m2
m2
freq=69.00GHz
freq (1.000GHz to 80.00GHz)
freq (1.000GHz to 80.00GHz)
Rev. A
IT-CDT 2016-18
Page 21 of 22
Observatorio de Yebes
Apartado 148, 19080 Guadalajara, SPAIN
YebeS
Effect of Source Bonding Wires in HEMT devices:
Probe Station Measurements
IAF 150 (chip height: 50 um) 1 BW each side
Smodel(1,2)
Smeas(1,2)
Smodel(1,1)
Smeas(1,1)
m1
m1
freq=46.00GHz
-1.5
freq (1.000GHz to 80.00GHz)
-1.0
-0.5
0.0
0.5
1.0
1.5
freq (1.000GHz to 80.00GHz)
DDS_File_Name
-8
-6
-4
-2
0
2
4
6
8
Smeas(2,2)
Smodel(2,2)
Smeas(2,1)
Smodel(2,1)
IAF_150_1BW_test
m2
m2
freq=39.00GHz
freq (1.000GHz to 80.00GHz)
freq (1.000GHz to 80.00GHz)
Rev. A
IT-CDT 2016-18
Page 22 of 22