Transmission Modes Different ways of characterizing the transmission.
Evaluation of Transmission Line Modeling Using Different ...
Transcript of Evaluation of Transmission Line Modeling Using Different ...
![Page 1: Evaluation of Transmission Line Modeling Using Different ...](https://reader035.fdocuments.us/reader035/viewer/2022081701/62de497f4e9b88128059eb62/html5/thumbnails/1.jpg)
Evaluation of Transmission Line Modeling Using Different De-
embedding Methods
Ryo Minami,Changyo Han,Kota MatsushitaKenichi Okada,and Akira Matsuzawa
Tokyo Institute of Technology, Japan
![Page 2: Evaluation of Transmission Line Modeling Using Different ...](https://reader035.fdocuments.us/reader035/viewer/2022081701/62de497f4e9b88128059eb62/html5/thumbnails/2.jpg)
2Outline• Background• About de-embedding• Different de-embedding methods
– Open-short method– Thru-only method– L-2L method
• Comparison of measured transmission line data
• Performance comparison of the PAs using different TL models
• Conclusion
![Page 3: Evaluation of Transmission Line Modeling Using Different ...](https://reader035.fdocuments.us/reader035/viewer/2022081701/62de497f4e9b88128059eb62/html5/thumbnails/3.jpg)
3
Atte
nuat
ion
[dB
/km
]
Frequency [GHz]
Background
IEEE 802.15.3c・QPSK ⇒ 3.5Gbps/ch・16QAM ⇒ 7.0Gbps/ch
[1] Rec. ITU-R P.676-2, Feb. 1997Advantage of 60GHzEnable communicationdistance is short.Be resistant to interfere
Suitable for short rangewireless Communications
Wireless Transmission of uncompressed HDTV
Gbps Wireless Communication
![Page 4: Evaluation of Transmission Line Modeling Using Different ...](https://reader035.fdocuments.us/reader035/viewer/2022081701/62de497f4e9b88128059eb62/html5/thumbnails/4.jpg)
4Transmission Line
capacitance(0.999 [µF])
inductance(1 [H]) inductance(1 [H])
capacitance(0.999 [µF])
The structure of TL.
Modeland photo.
PDK are only guaranteed to 20 GHz.The accurate models for 60 GHz are required.
TL is used for considering the circuit by distributedconstant.
DCAC
![Page 5: Evaluation of Transmission Line Modeling Using Different ...](https://reader035.fdocuments.us/reader035/viewer/2022081701/62de497f4e9b88128059eb62/html5/thumbnails/5.jpg)
5What is De-Embedding?
• Measurement data include parasitic components. • Parasitic components are not ignorable at 60GHz.
These are required to be removed.• Open-short method• Thru-only method• L-2L method
![Page 6: Evaluation of Transmission Line Modeling Using Different ...](https://reader035.fdocuments.us/reader035/viewer/2022081701/62de497f4e9b88128059eb62/html5/thumbnails/6.jpg)
6Open-short and thru-only methodOpen-short
Thru-only
Ideal short pattern cannot be achieved.
Affected by the thru length and probe coupling.
[2] H. Ito, et al, IMS 2008.
![Page 7: Evaluation of Transmission Line Modeling Using Different ...](https://reader035.fdocuments.us/reader035/viewer/2022081701/62de497f4e9b88128059eb62/html5/thumbnails/7.jpg)
7L-2L De-embedding method
TLTL
CanceledCanceled
× ×
T2L-1
TL
=
TPTS
Realize ideal thru pattern. Good isolation between probes. [3] N. Li, et al, IEICE
TRANS 2010.
![Page 8: Evaluation of Transmission Line Modeling Using Different ...](https://reader035.fdocuments.us/reader035/viewer/2022081701/62de497f4e9b88128059eb62/html5/thumbnails/8.jpg)
8Parameters of TL (α, β, Q, Z0)Z Z+∆ZTL
Equivalent circuit
Wave equation
![Page 9: Evaluation of Transmission Line Modeling Using Different ...](https://reader035.fdocuments.us/reader035/viewer/2022081701/62de497f4e9b88128059eb62/html5/thumbnails/9.jpg)
9Evaluation• open-short 200 m
300 m
400 m
• thru-only• L-2L
300 mPAD PAD
400 mPAD PAD
200 mPAD PAD
Make the models of TL using measurement result of different de-embedding methods.
These parameters must be the same regardless of the length of TLs.
Compare the parameter of
andunit length.
per
![Page 10: Evaluation of Transmission Line Modeling Using Different ...](https://reader035.fdocuments.us/reader035/viewer/2022081701/62de497f4e9b88128059eb62/html5/thumbnails/10.jpg)
10Evaluation, cont.Simulate the effect of different models by applying to a 4-stage PA.
RFin
RFout
Vgs1
Vds1 Vgs2
Vds2 Vgs3
Vds3 Vgs4
Vds4
160μm
270μm
70μm
60μm
370μm 130μm
70μm
170μm 270μm
20μm
140μm 20μm
370μm 290μm
50μm
310μm
150μm
70μm
W=40μm
W=40μm
W=80μm
W=80μm
100fF
100fF
100fF
100fF
75fF
4.25pF
9.875pF
8.75pF
6.875pF 5.75pF
5.75pF 5.75pF
7.25pF
The evaluation is focused on the accuracy of transmission line models, so the same models are used for transistor, MIM capacitor, etc.
Circuit schematic of 4-stage PA for simulation• common source topology• MIM capacitor is used for decoupling• 1.2V supply, 65nm CMOS
![Page 11: Evaluation of Transmission Line Modeling Using Different ...](https://reader035.fdocuments.us/reader035/viewer/2022081701/62de497f4e9b88128059eb62/html5/thumbnails/11.jpg)
11
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0 10 20 30 40 50 60 70
α [d
B/m
m]
Frequency [GHz]
4045505560657075808590
0 10 20 30 40 50 60 70
Z 0[Ω
]
Frequency [GHz]
0
5
10
15
20
25
30
35
40
0 10 20 30 40 50 60 70
Q
Frequency [GHz]
Measurement results using open-short method
Open-short method cannot be used for modeling.
0
500
1000
1500
2000
2500
3000
0 10 20 30 40 50 60 70
β [r
ad/m
]
Frequency [GHz]
model400 µm300 µm200 µm
model400 µm300 µm200 µm
model400 µm300 µm200 µm
model400 µm300 µm200 µm
![Page 12: Evaluation of Transmission Line Modeling Using Different ...](https://reader035.fdocuments.us/reader035/viewer/2022081701/62de497f4e9b88128059eb62/html5/thumbnails/12.jpg)
12
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0 10 20 30 40 50 60 70
α [d
B/m
m]
Frequency [GHz]
Measurement results using thru-only method
Z0 has 3 Ω difference at 60 GHz.
0
500
1000
1500
2000
2500
3000
0 10 20 30 40 50 60 70
β [r
ad/m
]
Frequency [GHz]
model400 µm300 µm200 µm
0
5
10
15
20
25
30
35
40
0 10 20 30 40 50 60 70
Q
Frequency [GHz]
4045505560657075808590
0 10 20 30 40 50 60 70
Z 0[Ω
]
Frequency [GHz]
model400 µm300 µm200 µm
model400 µm300 µm200 µm
model400 µm300 µm200 µm
![Page 13: Evaluation of Transmission Line Modeling Using Different ...](https://reader035.fdocuments.us/reader035/viewer/2022081701/62de497f4e9b88128059eb62/html5/thumbnails/13.jpg)
13
4045505560657075808590
0 10 20 30 40 50 60 70
Z o[Ω
]
Frequency [GHz]
0
5
10
15
20
25
30
35
40
0 10 20 30 40 50 60 70
Q
Frequency [GHz]
0
500
1000
1500
2000
2500
3000
0 10 20 30 40 50 60 70
β [r
ad/m
]
Frequency [GHz]
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0 10 20 30 40 50 60 70
α [d
B/m
m]
Frequency [GHz]
Measurement results using L-2L method
Z0 has 0.9 Ω difference at 60 GHz.
model400 µm300 µm200 µm
model400 µm300 µm200 µm
model400 µm300 µm200 µm
model400 µm300 µm200 µm
![Page 14: Evaluation of Transmission Line Modeling Using Different ...](https://reader035.fdocuments.us/reader035/viewer/2022081701/62de497f4e9b88128059eb62/html5/thumbnails/14.jpg)
14De-embedding error at 60GHz
Open-short Thru-only L-2L∆ α [%] 20.4 2.6∆ β [%] 32.7 5.7 0.1∆ Q [%] 21 1.5∆ Z0 [%] 15.3 6.1 1.7
This table shows the difference of transmission line parameters between different length of TLs.
L-2L method can obtain the smallest error of transmission line parameters.
![Page 15: Evaluation of Transmission Line Modeling Using Different ...](https://reader035.fdocuments.us/reader035/viewer/2022081701/62de497f4e9b88128059eb62/html5/thumbnails/15.jpg)
15The effect of different de-embedding method
Thru-only L-2LFreq. [GHz]
61.4 58.8
Gain [dB]
19.8 21.3
P1dB[dBm]
7.41 8.78
Psat[dBm]
11.0 11.4
Peak PAE [%]
6.51 7.45
Power [mW]
169 170
There are 1.5-dB gain error and 2.6-GHz frequency shift.Sim. error heavily depends on the de-embedding method.
Simulation result
0
5
10
15
20
25
30 40 50 60 70 80
S 21 [d
B]
Frequency [GHz]
Thru‐only
L‐2L
![Page 16: Evaluation of Transmission Line Modeling Using Different ...](https://reader035.fdocuments.us/reader035/viewer/2022081701/62de497f4e9b88128059eb62/html5/thumbnails/16.jpg)
16Conclusion• This presentation shows an evaluation of de-
embedding methods for mmW transceiver design.
• The L-2L de-embedding method is the most accurate for transmission-line characterization.
• The frequency shift with 4-stage PA using different de-embedding methods is 2.6 GHz which is over a channel of IEEE802.15.3c.
• The effect of different de-embedding method cannot be negligible.