Full Board Design and Verification of RF Systems Based on the Three-Dimensional Field Analysis KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University Advised by Professor SungWoo Hwang
Outlines • • • •
Motivation Introduction Full Board Simulation Design and Verification of the RF Systems – CMOS RF IC – RF LPA – RF PDP
• Conclusions
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
2/97
I-1. Motivation
Motivation I • Multimedia Services and Products I
CPU Clock Speed [Hz]
– Intel® Microprocessor Clock Speed
10G 1G 100M 10M CPU Clock Speed [Hz] Doubling time = 2.7 Years
1M 1970
1980
1990
2000
2010
2020
Year KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
4/97
Motivation II • Multimedia Services and Products II – Evolutions of the Communication Standards
High speed
Medium
2G/2.5G Cellular/PCS (0.8/1.8GHz)
4G (? GHz)
3G EVDO/ WCDMA (0.8/1.8/2.1GHz)
1G AMPS (0.8GHz)
PI (2.3 GHz)
Pedestrian
Y2010 WLAN (5.x GHz)
WLAN (2.4 GHz) Y1995
Y2005 Y2000
Fixed
Data rate Voice
Data
Image
Moving pictures
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Multimedia
Dec. 2004
5/97
Motivation III • The Requirements
1. High Operating Frequency 2. High Carrier Frequency 3. (High Level of Integration) Components and Systems
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
6/97
I-2. Introduction
Introduction I • Characteristics of the RF Components and Systems I Increased Guided Frequency Full Wave Analysis Electromagnetic Fields →
→
∇ × E = − jω B →
→
(Quasi) – TEM Analysis →
∇ × H = J + jω D →
∇ ⋅ D = ρv →
∇⋅B = 0
Distributed Circuits
dV = (R + jωL ) ⋅ I dz dI − = (G + jωC ) ⋅V dz −
KVL & KCL Analysis (SPICE) Lumped Circuits
V = jωLI I = jωCV V = IR
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
8/97
Introduction II • Characteristics of the RF Components and Systems II – Full Wave Analysis of the Complex RF System: Practicable ?
RF System Shielding Box Passive Components Pads and Bondwires Package MMIC Die SMA Connector KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
9/97
Introduction II • Design the RF Components and Systems I
Theory-based Design and Verification
Electromagnetic Fields Effects
Technique-based Design and Verification KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
10/97
Introduction III • Design the RF Components and Systems II 1. Theory-based Design and Verification
Z0 = E/H
Z0 = ?
• Accurate, But Hard to Apply to the Complex Structures KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
11/97
Introduction IV • Design the RF Components and Systems III 2. Technique-based Design and Verification Trial and Error
Know-how
=0.5pF
Signal Ground
=1.0pF
• Case-by-Case KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
12/97
Introduction V • Design the RF Components and Systems IV 3. Computer Aided Design (CAD) and Verification Theory-based Design and Verification
Electromagnetic Fields Effects
Computer Aided Design and Verification
Technique-based Design and Verification KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
13/97
Introduction VI • Computer Aided Design and Verification – Objectives of the Simulation: Measurements (RF)
Accurately Predicting and Describing the Measurement Results Simulation of the RF Component and System
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
14/97
II-1. Full Board Simulation – Introduction
Full Board Simulation – Introduction I • Type of the Simulator
– Solving KVL, KCL and Node Equations • Circuit-Level Simulator – KVL and KCL: Simplified Maxwell’s Equation • System-Level Simulator – Behavioral Model
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
General Method
• 2D/ 3D Device Simulator • 2D/ 3D Field Simulator
Dec. 2004
Complexity of Problem
– (Numerically) Solving Maxwell’s Equations
16/97
Full Board Simulation – Introduction II • Full Board Simulation Using Only the 3D Field and Device Simulators: – No Problems ? Measurements (RF)
Accurately Predicting and Describing the Measurement Results (Numerical) Field and Device Simulation KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
17/97
Full Board Simulation – Introduction III • Problem I: Large Size Problem – 60 inch RF PDP Panel
• Large Computational Resource • Large Simulation Time
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
18/97
Full Board Simulation – Introduction IV • Problem II: Simulation of the Full Board – 2.4 GHz Wireless LAN Card
• Full Board Simulation – Device Simulation – Field Simulation
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
19/97
Full Board Simulation – Introduction V • Full Board Simulation of the RF System I RF System, Governed by the Maxwell’s Equations
Segmentation I (Materials)
Conductor and Dielectrics
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Conductor, Dielectrics, and Semiconductors
Dec. 2004
20/97
Full Board Simulation – Introduction VI • Full Board Simulation of the RF System II Segmentation II (Problem Regions)
Segment #3b: Drain bias network and Transistor launcher
Segment #1: SMA-MSL (w. transition model) Segment #3a: Gate bias network and Transistor launcher
Segment #2: Uniform MSL
Segment #2: Uniform MSL
Segment #1: SMA-MSL (w. transition model)
Field Simulation S10P SNP16 File="G:\CircuitDesign_PowerGarden\ADS\Simulation libraries\20040711_mrf21030v1p1_parameter_extraction\red_multiport_segsetb_2.s10p"
L L55 L= 0.5 nH R=0.0
Term Term7 Num= 7 Z=50 Ohm
L L46 L=0.4 nH R=0.3
L L52 L=0.29 nH R=0.325
C C30 C=11 pF
Compact Circuit Modeling
R C R8 C35 R= 10 Ohm C= 2.2 uF
L L51 TLINP TLINP L=0.29 nH TL8 TL9 R=0.325 S2P Z=51.21313-1.26392e-10*freq+1.14622e-19*freq*freq-1.76927e-29*freq*freq*freq+ Z=51.21313-1.26392e-10*freq+1.14622e-19*freq*freq-1.76927e-29*freq*freq*freq+1.19269e-39*freq*freq*freq*freq-2.699e-50*freq*freq*freq*freq*freq 1.19269e-39*freq*freq*freq*freq-2.699e-50*freq*freq*freq*freq*freq SNP14 L=4.82 mm L=9.21 mm L File="G:\CircuitDesign_PowerGarden\040903_Test_board_modeling\SimulationData_HFSS\tran_msl_sma_left_10p0mm.s2p" K=3.13+1.51545e-11*freq K=3.13+ 1.51545e-11*freq L53 A=0.8 A=0.8 L=39.3 nH F=1 GHz F=1 GHz R=0.325 TanD=0.018 TanD=0.018 1
TLINP TL7 Z=50.65 Ohm L=14 mm K=1.672 A=0.001 F=1 GHz TanD=0.001
C C36 C=1.2 pF
Device Simulation
C C34 C=30 pF
2
Ref
1
10
9
8
2
7
3
6
4
L L49 L=0.29 nH R=0.325
C C33 C=5 pF
L L48 L=0.29 nH R=0.325
C C32 C=51 pF
L L47 L=0.29 nH R=0.325
C C31 C=1 nF
5
Ref
4 1
2
L L50 L L=0.29 nH L54 S4P R=0.325 SNP17 L= 0.01 nH File="G:\CircuitDesign_PowerGarden\040903_Test_board_modeling\SimulationData_HFSS\rem_red_multiport_segsetb_1.s4p" R=0.4 3
Ref
R R7 R=22 Ohm
1
TLINP TL3 Z=50.65 Ohm L=14 mm K=1.672 A=0.001 F=1 GHz TanD=0.001
2 Ref
S2P SNP13 File=
TLINP TL5 Z= L=4.82 mm K=3.13+1.51545e-11*freq A=0.8 F=1 GHz TanD=0.018
Term Term8 Num=8 Z=50 Ohm
Full Board Simulation of the RF System Dec. 2004
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
21/97
Full Board Simulation – Introduction VII
n tio ca ifi er V
ra ct io rE xt Pa ra m et e
d an
(S im
ul
at
io
n
Measurements (RF)
n ig es eD at ur cc A
n)
• Full Board Simulation of the RF System III
(Numerical) Field and Device Simulation
Simulation of the RF Component and System
Segmentation and Compact Circuit Modeling KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
22/97
II-2. Full Board Simulation – Segmentation
Full Board Simulation – Segmentation I • Segmentation Method: Transition Regions I – Expected Causes of the Segmentation Error n ion o i g e g e R Z2 on R i t i s n a Tr gion e R ave Z1 W d itte m s n a Tr ave W t en Incid a ve W d e ct Refle
Impedance Discontinuity Propagating Mode Reflected Wave
Geometric Discontinuity Non-Propagating Mode Capacitive Fringing Current Crowding
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
24/97
Full Board Simulation – Segmentation II • Segmentation Method: Transition Regions II – Types of the Simulated Structure Type-I: Impedance Discontinuity Different Z0, But Same Geometric Parameters
Type-II: Geometric Discontinuity Same Z0, But Different Geometric Parameters
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
25/97
Full Board Simulation – Segmentation III • Segmentation Method: Transition Regions III – Types of the Simulation Method
Full 3D Field Simulation Reference
Full Board Simulation I (Problem Regions) Uniform + Transition + Uniform Composition
Full Board Simulation II (Materials) Uniform + Uniform Composition KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
26/97
Full Board Simulation – Segmentation IV • Segmentation Method: Transition Regions IV – Simulation Results: Type-I 0.5mm-0.5mm: 0.6mm-0.6mm: 0.7mm-0.7mm: 0.8mm-0.8mm: 0.9mm-0.9mm: 1.0mm-1.0mm:
3.0 2.5 2.0
3D-F.B.Sim#1 3D-F.B.Sim#1 3D-F.B.Sim#1 3D-F.B.Sim#1 3D-F.B.Sim#1 3D-F.B.Sim#1
3D-F.B.Sim#2 3D-F.B.Sim#2 3D-F.B.Sim#2 3D-F.B.Sim#2 3D-F.B.Sim#2 3D-F.B.Sim#2
1.5 1.0 0.5 0.0 -0.5 -1.0 0
5
10
15
20
S21_3D - S21_F.B.Sim. [dB]
S11_3D - S11_F.B.Sim. [dB]
• 60Ω to 30Ω Transition 1.0 0.5mm-0.5mm: 0.6mm-0.6mm: 0.7mm-0.7mm: 0.8mm-0.8mm: 0.9mm-0.9mm: 1.0mm-1.0mm:
0.8 0.6 0.4
3D-F.B.Sim#1 3D-F.B.Sim#1 3D-F.B.Sim#1 3D-F.B.Sim#1 3D-F.B.Sim#1 3D-F.B.Sim#1
3D-F.B.Sim#2 3D-F.B.Sim#2 3D-F.B.Sim#2 3D-F.B.Sim#2 3D-F.B.Sim#2 3D-F.B.Sim#2
0.2 0.0 -0.2 0
5
10
15
Frequency [GHz]
Frequency [GHz]
S11 Response
S21 Response
20
• Impedance Steps Cause No Segmentation Error KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
27/97
Full Board Simulation – Segmentation V • Segmentation Method: Transition Regions V – Simulation Results: Type-II
1.0mm-0.6mm: 1.0mm-0.7mm: 1.0mm-0.8mm: 1.0mm-0.9mm:
15 10 5
3D-F.B.Sim#1 3D-F.B.Sim#1 3D-F.B.Sim#1 3D-F.B.Sim#1
3D-F.B.Sim#2 3D-F.B.Sim#2 3D-F.B.Sim#2 3D-F.B.Sim#2
0 -5 -10 -15 -20 -25 -30 0
5
10
15
20
S21_3D - S21_F.B.Sim. [dB]
S11_3D - S11_F.B.Sim. [dB]
• 30Ω to 30Ω Transition I 1.0mm-0.6mm: 1.0mm-0.7mm: 1.0mm-0.8mm: 1.0mm-0.9mm:
1.5
3D-F.B.Sim#1 3D-F.B.Sim#1 3D-F.B.Sim#1 3D-F.B.Sim#1
3D-F.B.Sim#2 3D-F.B.Sim#2 3D-F.B.Sim#2 3D-F.B.Sim#2
1.0 0.5 0.0 -0.5 0
Frequency [GHz]
5
10
15
20
Frequency [GHz]
S11 Response S21 Response • Geometric Discontinuity Cause the Segmentation Error KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
28/97
Full Board Simulation – Segmentation VI • Segmentation Method: Transition Regions VI – Simulation Results: Type-II
7 6 5 4 3
[dB]
8
1.2
| RMS error of |S21
RMS error of |S11
| [dB]
• 30Ω to 30Ω Transition II 1.0 0.8 0.6 0.4
2 0.2
1 0 (Ge 0.0 0.2 om etri 0.4 cal_ S te p)x (ε
r1
m] m / [1 8
6 4
-ε
r2
o ity c o el
2
0.6 0.5
0. 5
0.8 0
) [m m]
/v eq Fr
gh f li
RMS Error of the S11
t
0.0 (Ge 0.0 0.2 om etri 0.4 cal_ S te p) x (ε
r1
m] m [1 / 8
6
h lig of
4 2
0.6 0.5
-ε
r2
0.5
0.8 0
) [m m]
eq Fr
y cit o l /ve
t
RMS Error of the S21
• Electrically Small Geometric Discontinuity Cause No Segmentation Error KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
29/97
Full Board Simulation – Segmentation VII • Segmentation Method: Conclusion – Impedance Discontinuity Only • Cause No Segmentation Error – With Transition Model: Segmentation II (Problem Regions) – Without Transition Model: Segmentation I (Materials)
– Geometric Discontinuity Only • Cause No Segmentation Error – With Transition Model
• Cause the Segmentation Error – Without Transition Model
• Electrically Small Geometric Discontinuity Cause No Segmentation Error
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
30/97
II-3. Full Board Simulation – DUT Model Extraction
Full Board Simulation – DUT Model Extraction I • DUT (Device Under Test) Model Extraction I – Problems • To Apply Coaxial Instruments to Non-Coaxial DUT
– An Ideal Test Fixture • Provide the Electrical Transparent Connections Between the Instruments and the DUT
Cascade® Microtech
Inter-Continental Microwave
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
32/97
Full Board Simulation – DUT model Extraction II • DUT Model Extraction II – Method I: Direct Measurements • Response Calibration: S21 Only • Full Two-port Calibration: SOLT and TRL
Characterization of the Test Fixture: (Calibration Standards: OPEN, LINE, SHORT,LOAD, …,) DUT with Test Fixture
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
33/97
Full Board Simulation – DUT Model Extraction III • DUT Model Extraction III – Method II: Modeling • Port Extension: Constant Z0 and γ • De-Embedding: Full Fixture Modeling
Modeling of the Test Fixture: No Calibration Standards DUT with Test Fixture
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
34/97
Full Board Simulation – DUT Model Extraction IV • DUT Model Extraction IV – De-Embedding
[S]_Left [S]_DUT
[S]_Right
[S]_DUT
S DUT ,11 = [ S ]_ DUT = ⎛ S DUT ,11 S DUT ,12 ⎞ ⎟ ⎜ ⎜ S DUT ,21 S DUT ,22 ⎟ = ⎠ ⎝
S Full ,12 S Full , 21S Right ,11 + ( S Full ,11 − S Left ,11 )( S Full , 22 S Right ,11 + S Right ,12 S Right ,12 − S Right ,11S Right , 22 )
S DUT ,12 =
S Full ,12 S Left ,12 S Right ,12 D
D S DUT , 21 =
S Full , 21S Left , 21S Right , 21 D
S DUT , 22 =
− S Full ,12 S Full , 21 S Left , 22 + ( S Full , 22 − S Right , 22 )( S Left ,12 S Left ,12 + S Full ,11 S Left , 22 − S Left ,11 S Left , 22 ) D
D = − S Full ,12 S Full , 21S Left , 22 S Right ,11 + S Full , 22 ( S Left ,12 S Left ,12 + ( S Full ,11 − S Left ,11 ) S Left , 22 ) S Right ,11 + ( S Left ,12 S Left ,12 + S Full ,11S Left , 22 − S Left ,11S Left , 22 )( S Right ,12 S Right ,12 − S Right ,11S Right , 22 ) KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
35/97
Full Board Simulation – DUT Model Extraction V • DUT Model Extraction V – (Internal Port) Impedance Analysis Direct Measurements: Requires Coaxial Interfaces Netwo rk Ana ly zer
Modeling: Can be Performed at Any Segments KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
36/97
II-4. Full Board Simulation – Test Fixture Model: Modeling
Full Board Simulation – Test Fixture model I • Modeling of the Test Fixture I – Simulation Parameter Extraction • SPICE Circuit Simulation gm, rds, μn, μp, λ, Cds, Cgs, Cgd,……,
Accurate I-V, C-V
• 3D Field Simulation Effective Dielectric Constant (εeff) Characteristic Impedance (Z0)
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Accurate E/H-Fields
Dec. 2004
38/97
Full Board Simulation – Test Fixture model II • Modeling of the Test Fixture II – Problems: Coupled RF Characteristics
Uniform SMA Region
Transition Uniform MSL Region Region
Extraction, Separately KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
39/97
Full Board Simulation – Test Fixture model III • Main Idea – The Characteristics of the Uniform Regions have Nothing to do with Those of Transition Regions
SSSM-type Test Jig
SSSM-type test jig
LSSM-type Test Jig
LSSM-type test jig
SSLM-type Test Jig
SSLM-type test jig
LSLM-type test jig LSLM-type Test Jig KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
40/97
Full Board Simulation – Test Fixture model IV • Simulation Parameter Extractions I – Effective Dielectric Constant (εr) Δl c Δt ⎞ 2 ⎛ v= = ⇒ ε eff = ⎜ c ⎟ Δt ε eff ⎝ Δl ⎠ SSSM-Type SSLM-Type
200
LSSM-Type LSLM-Type
150
S21 [Deg.]
100 50 0 -50 -100 -150 -200 0
1
2
3
4
5
Frequency [GHz]
VNA Measurements KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
41/97
Full Board Simulation – Test Fixture model V • Simulation Parameter Extractions II – Characteristic Impedance (Z0) 0.2 Uniform SMA section
Z (l ) − Z 0 ΓL (l ) = L Z L (l ) + Z 0
SMA-to-microstrip line transition
0.1 0.0
ΓρL(l)
-0.1 -0.2
TDR equipment
-0.3
Uniform microstrip line section (Difference in electrical length)
-0.4
ρ of LSSM-type ρ of LSLM-type
-0.5 -0.6 33.6
34.0
34.4
34.8
35.2
35.6
36.0
Time [ns]
TDR Measurements KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
42/97
Full Board Simulation – Test Fixture model VI • 3D Field Simulation – LSSM-type Test Jig 5mm 1.42mm 32mm 30mm
0.76mm
0.67mm 2.0mm 3.05mm
– LSLM-type Test Jig
60mm KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
43/97
Full Board Simulation – Test Fixture model VII • Compact Circuit Modeling I – Uniform Region of the Test Fixture • ADS TLINP Model Board-level interconnects SMA connector
L [mm] ZO [Ω]
(length of
K
the thread) 50.6
14/ 32
1.67
A (@1GHz)
Tan D
0
0.001
1.2
0.018
51.21313-1.26392⋅10-10× freq+1.14622⋅10-19×freq2Microstrip line
1.76927⋅10-29×
3.13+1.515 20/ 50
freq3+1.19269⋅10-39×freq4-
45⋅1011
×freq
2.699⋅10-50×freq5
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
44/97
Full Board Simulation – Test Fixture model VIII • Compact Circuit Modeling II – Non-uniform Region of the Test Fixture
[S ]transition = [eγ _ left⋅l1 ] ⋅ [S ]overall ⋅ [eγ _ right⋅l2 ] Uniform region, l1 (port1) Transition region (to be modeled) Uniform region, l2 (port2)
-
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
=
Dec. 2004
45/97
II-5. Full Board Simulation – Test Fixture Model: Verification
Full Board Simulation – Verification I • Measured-/ 3D Field, Circuit-Level Simulated Results I – SSSM-type Test Jig 200
-10
150
-20
S11 [Deg.]
S11 [dB]
100 -30 -40 -50 -60
VNA_Measurement_|S11| 3D_Field_Simulation_|S11| ADS_Simulation_|S11|
0 -50 -100 -150
-0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5
VNA_Measurement_ph(S11) 3D_Field_Simulation_ph(S11) -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 ADS_Simulation_ph(S11)
Frequency [GHz]
Frequency [GHz]
-70
0.1
-200
200
0.0
150
-0.1
100
S21 [dB]
-0.2
S21 [dB]
50
-0.3 -0.4 -0.5 -0.6 -0.7 -0.8
VNA_Measurement_|S21| 3D_Field_Simulation_|S21| ADS_Simulation_|S21|
-0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5
50 0 -50 -100 -150 -200
VNA_Measurement_ph(S21) 3D_Field_Simulation_ph(S21) -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 ADS_Simulation_ph(S11)
Frequency [GHz] KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Frequency [GHz] Dec. 2004
47/97
Full Board Simulation – Verification II • Measured-/ 3D Field, Circuit-Level Simulated Results II – LSSM-type Test Jig 200
-10
150
-20
S11 [Deg.]
S11 [dB]
100 -30 -40 -50 -60 -70
VNA_Measurement_|S11| 3D_Field_Simulation_|S11| ADS_Simulation_|S11|
50 0 -50 -100 -150
-0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5
VNA_Measurement_ph(S11) 3D_Field_Simulation_ph(S11) -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 ADS_Simulation_ph(S11)
Frequency [GHz]
Frequency [GHz]
-200
200
0.0
150 100
S21 [Deg.]
S21 [dB]
-0.2 -0.4 -0.6 -0.8
VNA_Measurement_|S21| 3D_Field_Simulation_|S21| ADS_Simulation_|S21|
-1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5
50 0 -50 -100 -150 VNA_Measurement_ph(S21) 3D_Field_Simulation_ph(S21) -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 ADS_Simulation_ph(S21)
-200
Frequency [GHz] KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Frequency [GHz] Dec. 2004
48/97
Full Board Simulation – Verification III • Measured-/ 3D Field, Circuit-Level Simulated Results III – SSLM-type Test Jig 200
-10
150 100
S11 [Deg.]
S11 [dB]
-20 -30 -40 -50 -60
VNA_Measurement_|S11| 3D_Field_Simulation_|S11| ADS_Simulation_|S11|
Frequency [GHz]
Frequency [GHz] 200 150 100
S21 [Deg.]
-0.4
S21 [dB]
-100 VNA_Measurement_ph(S11) 3D_Field_Simulation_ph(S11) -0.5 0.0 0.5 1.0 1.5 2.0 2.5ADS_Simulation_ph(S11) 3.0 3.5 4.0 4.5 5.0 5.5
-0.2 -0.6 -0.8 -1.0 -1.4
-50
-200
0.0
-1.2
0
-150
-70 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 0.2
50
VNA_Measurement_|S21| 3D_Field_Simulation_|S21| ADS_Simulation_|S21|
-1.6 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5
50 0 -50 -100 -150 VNA_Measurement_ph(S21) 3D_Field_Simulation_ph(S21) -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 ADS_Simulation_ph(S21)
-200
Frequency [GHz] KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Frequency [GHz] Dec. 2004
49/97
Full Board Simulation – Verification IV • Measured-/ 3D Field, Circuit-Level Simulated Results IV – LSLM-type Test Jig -10
200
-20
150 100
S11 [Deg.]
S11 [dB]
-30 -40 -50 -60 -70
VNA_Measurement_|S11| 3D_Field_Simulation_|S11| ADS_Simulation_|S11|
Frequency [GHz]
Frequency [GHz] 200 150
-0.6 -0.8 -1.0 VNA_Measurement_|S11| 3D_Field_Simulation_|S11| ADS_Simulation_|S11|
S21 [Deg.]
100
-0.4
S21 [dB]
-100
VNA_Measurement_ph(S11) -0.5 0.0 0.5 1.0 1.5 2.0 2.53D_Field_Simulation_ph(S11) 3.0 3.5 4.0 4.5 5.0 5.5 ADS_Simulation_ph(S11)
-0.2
-1.4
-50
-200
0.0
-1.2
0
-150
-80 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 0.2
50
50 0 -50 -100 -150 -200
-1.6 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5
VNA_Measurement_ph(S21) 3D_Field_Simulation_ph(S21) -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 ADS_Simulation_ph(S21)
Frequency [GHz] KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Frequency [GHz] Dec. 2004
50/97
II-6. Full Board Simulation – Test Fixture Model: Structure Optimization
Full Board Simulation – Structure Optimization I • Structure Optimization of the Test Fixture I – Design Idea
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
52/97
Full Board Simulation – Structure Optimization II • Structure Optimization of the Test Fixture II – Implementation
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
53/97
Full Board Simulation – Structure Optimization III • Structure Optimization of the Test Fixture III – Experimental Verification -10 -20
S11 [dB]
-30 -40 -50 -60
Original_measurement Optimized_measurement 3D_field_simulation
-70 -80 0
1
2
3
4
Original_measurement Optimized_measurement 3D_field_simulation
0.0
5
-0.2
S21 [dB]
Frequency [GHz]
-0.4 -0.6 -0.8 -1.0 0
1
2
3
4
5
Frequency [GHz] KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
54/97
III-1. Design and Verification of the RF Systems: Introduction
Design and Verification – Introduction • Case Studies: 1. Electrically Small, But Complex RF System CMOS RF IC 2. RF System, which Requires Identical RF Components RF LPA 3. Simple, But Repeated Large Size RF System RF PDP
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
56/97
III-2. Design and Verification of the RF Systems – CMOS RF IC
CMOS RF IC – Introduction I • RF ICs in CMOS (Year 2004) – Dominant • Relatively Relaxed Communication Standard • Wireless LAN and Bluetooth
– Making Inroads • GSM Cellular Transceivers and GPS Receiver
– Perspective • RF System-On-a-Chip in CMOS
• CAD Tools Play an Important Role in the Development of RF ICs in CMOS
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
58/97
CMOS RF IC – Introduction II • Problem Definition: CMOS RF IC – Plentiful Active Device Models • Use the SPICE Simulator
– But, Lack of the RF Models • • • • •
Test Fixture Package Bondwire Pad On-Chip Interconnects
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
59/97
CMOS RF IC – Full Board Simulation Using SPICE I
Bonding wires
Silicon IC's (Device models)
Board Interconnects Pads
Verification
• Obstacles to the Accurate SPICE Simulation Verification is Performed in Time- and Frequency-Domain - Time Domain Simulation in SPICE: OK - S-Parameter Simulation in SPICE: ?
Packaged
Design
Lead Frames
Full Board Simulation - Segmentation Method - Accurate RF Models: Compact Circuit Modeling
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
60/97
CMOS RF IC – Full Board Simulation Using SPICE II • Model Parameter Extraction I – Test Fixture
0.25 mm 0.254 mm
– Package
0.65 mm
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
61/97
CMOS RF IC – Full Board Simulation Using SPICE III • Model Parameter Extraction II – Bondwires 0.49 mm
0.025 mm 3.46 mm
– Pads
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
100 μm
Dec. 2004
20 μm
62/97
CMOS RF IC – Full Board Simulation Using SPICE IV • Segmentation: Verifications I – Test Fixture to Package Lead I • No Geometric Discontinuity Air Wave port (90Ω) 10mm
0.4mm 0.76mm 0.02mm
Lumped Gap port (50Ω) 0.4mm
Lumped Gap port 0.4mm (50Ω)
FR4 0.86mm
3mm
15mm 4.4mm
Paddle (connected to PCB GND) KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
63/97
CMOS RF IC – Full Board Simulation Using SPICE V • Segmentation: Verifications II -5
100
-10
50
S11 [Deg.]
S11 [dB]
– Test Fixture to Package Lead II
-15 -20 -25
Full 3D Field Simulation Segmenation without Transition Model
0 -50 Full 3D Field Simulation Segmenation without Transition Model
-100
-30 0
5
10
15
0
20
5
10
15
20
Frequency [GHz]
Frequency [GHz] 0
200 150 100
S21 [Deg.]
S21 [dB]
-1 -2 -3
10
15
Frequency [GHz]
-50 -100 Full 3D Field Simulation Segmenation without Transition Model
-200
-4 5
0
-150
Full 3D Field Simulation Segmenation without Transition Model 0
50
20
0
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
5
10
15
20
Frequency [GHz] Dec. 2004
64/97
CMOS RF IC – Full Board Simulation Using SPICE VI • Segmentation: Verifications III – Package Lead to Bondwire Transition I • Electrically Small Structure 0.4mm
Lumped Gap port (50Ω)
Wave port 0.4mm 0.18mm
0.03mm
1.12mm 0.67mm
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
65/97
CMOS RF IC – Full Board Simulation Using SPICE VII • Segmentation: Verifications IV -5 -10 -15 -20 -25 -30 -35 -40 -45 -50
80 60
S11 [Deg.]
S11 [dB]
– Package Lead to Bondwire Transition II
Full 3D Field Simulation Segmenation without Transition Model 0
5
10
15
40 20 Full 3D Field Simulation Segmenation without Transition Model
0
20
0
5
0.2 0.0 -0.2 -0.4 -0.6 -0.8 -1.0 -1.2 -1.4 -1.6 -1.8
15
20
Frequency [GHz] 20 0 -20
S21 [Deg.]
S21 [dB]
Frequency [GHz]
10
5
10
15
Frequency [GHz]
-60 -80 -100 -120
Full 3D Field Simulation Segmenation without Transition Model 0
-40
Full 3D Field Simulation Segmenation without Transition Model
-140 20
0
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
5
10
15
Frequency [GHz]
Dec. 2004
20
66/97
CMOS RF IC – Full Board Simulation Using SPICE VIII • Full Board Simulation I – Compact Circuit Modeling: Simulation Library Package Block (n)th (Lead frame+bonding wire +pad+substrate) LLS RLS LB L= Ln ZO=Zn LLM CLC L= Ln+1 ZO=Zn+1 L= L2n ZO=Z2n
CPC
LBM
CLS
CPS
CSUB RSUB
Package Block (n+1)th
Passive Components & Active Components
Package Block (2n)th
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
67/97
CMOS RF IC – Full Board Simulation Using SPICE IX • Full Board Simulation II – S-Parameter Simulation in SPICE Vi − Z 0 I i
• Equations
2 Z 0 Vi − Z 0 I i b Sij = i = = V Z I + aj V j + Z0 I j j 0 j 2 Z0
• Equivalent Circuits 50 Ω Port1
AC 0V
Port2
ii +
50 Ω AC 1V @ Operating Frequency
Vi ij
-
+
Vj
V = Sii
DC_bias 2Vi
AC 1
1Ω
Porti N-port network Port j V = Sij
-
PortN-1
2Vi
1Ω
PortN KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
68/97
CMOS RF IC – Experimental Verification • Measured- and Simulated Results – Down-converter for PCS Application 0 20
-5 -10
S11 [dB]
Voltage [mV]
10 0
-15
Measured Simulated
-20
10
-25
Measured IF Simulated IF
20 -200
-100
0
100
200
300
-30 400
1.4
1.6
1.8
2.0
2.2
Time [ns]
Frequency [GHz]
Time Domain
Frequency Domain
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
2.4
2.6
69/97
III-3. Design and Verification of the RF Systems – RF Linearized Power Amplifier (LPA)
RF LPA – Introduction I • The Relationship: Rated Output Power Level, Linearity Specification and Efficiency Output Power Level Efficiency Linearity Linearity Specification Unlinearized PA
PA with Back-Off
Linearized PA
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
71/97
RF LPA – Introduction II • Linearization of the RF Power Amplifier – Schematics of the IMD Cancellation Approach Applied RF signal @ t1 [s]
Known IMD(@t1) @t2 [s]
Nonlinear device (e.g. Amplifier, Mixer, etc.)
IMD(@t1) linearization @ t2 [s] Requires output delay time t2-t1 [s]
Linearizer (cancellation)
t
t
t
– Schematics of the IMD Prediction Approach Applied RF signal @ t1 [s]
Predicted IMD(@t1) @t2 [s]
Linearizer (prediction)
t
IMD(@t1) linearization @ t2 [s] Requires no output delay time
Nonlinear device (e.g. Amplifier, Mixer, etc.)
t
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
t
Dec. 2004
72/97
RF LPA – Introduction III • Efficiency Degradation Factors – Output Delay Line Diameter Delay/Frequency 0.047’’ 0.0865’’ 0.141’’ 3.30 1.78 0.99 1GHz 12ns 4.23 2.56 1.46 2GHz 4.12 2.22 1.24 1GHz 15ns 5.29 3.20 1.82 2GHz Delay Line Loss in [dB]
0.25’’ 0.62 0.97 0.77 1.22
– Output Power Combiner Coupling [dB]
6
10
15
20
30
Theoretical Minimum Main 1.20 0.46 0.14 0.04 0.004 Line Insertion Loss [dB]
– High Linear Error Amplifier • Trade-off: Output Power Combiner KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
73/97
RF LPA – Proposed Analog Predistortion-type LPA I • IMD Prediction Approaches: Predistortion-type LPA – Conventional Predistortion-type LPA Pi,PD Input Pi,LPA
Predistorter
Po,PA
Po,PD Pi,PA P3rd,PD
P3rd,PA
Output Po,LPA P3rd,LPA
Drive Amplifier Main Amplifier
– Proposed Predistortion-type LPA Pi,PD Input Pi,LPA
Po,PD Pi,PA
Po,PA
P3rd,PD
P3rd,PA
Drive Amplifier Predistorter
Output Po,LPA P3rd,LPA
Main Amplifier
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
74/97
RF LPA – Proposed Analog Predistortion-type LPA II • Block Diagram of the Proposed LPA System Note: P 3rd = 3P out – 2OIP 3 : Positive comp. : Negative comp.
Identical devices & Identical input level (@fundamental freq.)
Main Amplifier
a+G1
a
a+G1+6
3(a+G1)-2A-G1
3(a+G1)-2A
Tr.
3(a+G1)-2A Pout=a+G1
Tr.
4 way
a
Tr.
P3rd=3(a+G1)-2A
Tr.
(a+3) dBm
2 way
Tr. Path I
Drive Amplifier G=G 1 dB OIP3=A dBm
CPL
Delay
4 way
2 way
Main Amplifier G=G 1 dB OIP3=A+10log(4) dBm
ATT 2•(3(a+G1)-2A)
Path II
a
Delay
Vector Modulator
2 way
Vector Modulator
Error Amp.
Predistorter & Drive Amplifier KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
75/97
RF LPA – Design the Tr. Blocks I • Test Fixture Modeling I – Full Board Simulation Model: Segmentation Segment #3b: Drain bias network and Transistor launcher Segment #2: Uniform MSL
Segment #1: SMA-MSL (w. transition model) Segment #3a: Gate bias network and Transistor launcher
Segment #1: SMA-MSL (w. transition model)
Segment #2: Uniform MSL
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
76/97
RF LPA – Design the Tr. Blocks II • Test Fixture Modeling II – Measured- and Simulated- Results -5
Measured |S21| Simulated |S21|
|S21| [dB]
-6
1.0j 2.0j
0.5j
-7 -8 -9
5.0j
0.2j
-10 1.8
1.9
2.0
2.1
2.2
2.3
2.4
Frequency [GHz] 0.2
0.5
1.0
2.0
5.0
-0.2j
-2.0j
-0.5j
Measured |S11| Simulated |S11| Measured |S22| -5.0j Simulated |S22| From 1.8GHz to 2.4 GHz (Normalized to 50 Ω)
-1.0j KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
77/97
RF LPA – Design the Tr. Blocks III • Experimental Verification I – S11: Conjugate Impedance Matching 1.0j 0.5j
2.0j
0.2j
5.0j
0.2
0.5
1.0
2.0
5.0
-0.2j
-5.0j
-0.5j
-2.0j -1.0j
Measured |S11|_Tr.#0 Simulated |S11|_Tr.#0 Measured |S11|_Tr.#1 Simulated |S11|_Tr.#1 Measured |S11|_Tr.#2 Simulated |S11|_Tr.#2 Measured |S11|_Tr.#3 Simulated |S11|_Tr.#3 Measured |S11|_Tr.#4 Simulated |S11|_Tr.#4 From 2.0GHz to 2.3 GHz (Normalized to 50 Ω)
RMS modeling errors of |S11|, in [dB] Tr.#0 Tr.#1 Tr.#2 Tr.#3 Tr.#4 1.934 1.704 1.949 1.155 0.663 Deviated characteristics of each Tr. from those of Tr.#0 Tr.#0 Tr.#1 Tr.#2 Tr.#3 Tr.#4 1.702 2.413 1.032 0.843
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
78/97
RF LPA – Design the Tr. Blocks IV • Experimental Verification II – S22: Conjugate Impedance Matching 1.0j 0.5j
2.0j
0.2j
5.0j
0.2
0.5
1.0
2.0
5.0
-0.2j
-5.0j
-0.5j
Measured |S22|_Tr.#0 Simulated |S22|_Tr.#0 Measured |S22|_Tr.#1 Simulated |S22|_Tr.#1 Measured |S22|_Tr.#2 Simulated |S22|_Tr.#2 Measured |S22|_Tr.#3 Simulated |S22|_Tr.#3 Measured |S22|_Tr.#4 Simulated |S22|_Tr.#4 From 2.0GHz to 2.3 GHz (Normalized to 50 Ω)
-2.0j -1.0j
RMS modeling errors of |S22|, in [dB] Tr.#0 Tr.#1 Tr.#2 Tr.#3 Tr.#4 1.426 1.971 2.112 3.868 0.167 Deviated characteristics of each Tr. from those of Tr.#0 Tr.#0 Tr.#1 Tr.#2 Tr.#3 Tr.#4 1.491 1.746 4.767 1.387
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
79/97
RF LPA – Design the Tr. Blocks V • Experimental Verification III – S21: Conjugate Impedance Matching Measured |S21|_Tr.#0 Measured |S21|_Tr.#1 Measured |S21|_Tr.#2 Measured |S21|_Tr.#3 Measured |S21|_Tr.#4
16 15
S21 [dB]
14
Simulated |S21|_Tr.#0 Simulated |S21|_Tr.#1 Simulated |S21|_Tr.#2 Simulated |S21|_Tr.#3 Simulated |S21|_Tr.#4
13 12 11 10 9 2.00
2.05
2.10
2.15
Frequency [GHz]
2.20
2.25
2.30
RMS modeling errors of |S21|, in [dB] Tr.#0 Tr.#1 Tr.#2 Tr.#3 Tr.#4 0.326 0.305 0.294 0.267 0.128 Deviated characteristics of each Tr. from those of Tr.#0 Tr.#0 Tr.#1 Tr.#2 Tr.#3 Tr.#4 0.274 0.311 0.435 0.250
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
80/97
RF LPA – Implementation of the RF LPA System I • Proposed Predistortion-type LPA System
Main Amplifier
Control part
Drive Amplifier 4way Vector modulator ATT Error amp. CPL 2way
Delay
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
81/97
RF LPA – Implementation of the RF LPA System II • Experimental Set-up Power meter
Spectrum analyzer High power attenuator Signal generator Power supply
DUT
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
82/97
RF LPA – Experiments I • Performances I: @ Specific Power – Test Conditions • Pout = 36dBm • Two-tone Signal: fc = 2150 MHz, Δf = 1MHz • WCDMA Signal: fc = 2150 MHz, 1FA Test Model-I 64 DPCH Unlinearized Response
Linearized Response KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
83/97
RF LPA – Experiments II • Performances II: Output Dynamic Ranges – Test Conditions • Two-tone Signal: fc = 2150 MHz, Δf = 1MHz • WCDMA Signal: fc = 2150 MHz, 1FA Test Model-I 64 DPCH
– Improvements -35
Efficiency_unlinearized Efficiency_linearized
IMD3 [dBc]
-40 -45
17dB
-50
Linearity Specification
-55 -60 -65 28
30
32
34
36
38
40
-40
Unlinearized ACLR (@5MHz offset) Linearized ACLR (@5MHz offset)
-45
Efficiency [%]
12 11 10 9 8 7 6 5 4 3 2 1
ACLR [dB]
IMD3_ IMD3_
-50
Linearity Specification
-55
13dB
-60 -65 30
Pout [dBm] KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
32
34
36
38
Pout [dBm] Dec. 2004
84/97
III-4. Design and Verification of the RF Systems – RF PDP
RF PDP – Introduction I • Development of Large Size Flat Panel Displays – The Plasma Display Panel (PDP) • Most Feasible Technologies due to – Cost-effective Manufacturability – Wide Viewing Angle – Light Weight and Thinness
• But, Large DC Power Consumption
– RF Signal Biasing • The Plasma Excitation Can Occur at much Lower Electric Field • Leads to Higher Power Efficiency (≅ 3 Times) KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
86/97
RF PDP – Introduction II • Problem Definition: RF PDP – Simple Structure – Large Size RF System • Repeated Structures of the Element Pixels
– Electrical Length of the DUT Physical size
AC PDP (conventional) More than 1 m
RF PDP (newly developed) More than 1 m
Operating frequency
A few tens kHz
A few tens MHz
The order of v⋅L The order or phase difference (Φ ) : the same v assumed
104
107
MMICs (conventional) A few mm A few ~ A few tens GHz 106 ~ 107
10-3
1
10-1 – 1
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
87/97
RF PDP – Full Board Simulation of the RF PDP I • Segmentation: Full Board Simulation Model
(Segmented) Full Board Simulation Model 4x4 cell block
4x4 cell block
4x4 cell block
4x4 cell block
28 line connector
3D Field Simulation Model LC RC CC N type connector
28 line connector 4x4 cell block
4x4 cell block
4x4 cell block
4x4 cell block
4x4 cell block
4x4 cell block
4x4 cell block
4x4 cell block
28 line connector
28 line connector 4x4 cell block
4x4 cell block
CP1
4x4 cell block
4x4 cell block CP2
ground frame RG
LG
RG
LG
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
RG
LG
Dec. 2004
RG
LG
88/97
RF PDP – Full Board Simulation of the RF PDP II • DUT Model Electrical Parameter Extraction I – RF PDP Element Pixels I
1.00 mm
1.26 mm 1.26 mm
Front View
3D Field Simulation Model
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
89/97
RF PDP – Full Board Simulation of the RF PDP III • DUT Model Electrical Parameter Extraction II – RF PDP Element Pixels II 4x4 cell block sustain line1
5.04 mm
KS1
5.04 mm
LSUS
sus_line 2,3,4 KC1 KS2 sus_line 3,4 KS3 sus_line 4
LSCAN
scan_line 2,3,4 KC2 scan_line 3,4 KC3
2.80 mm
KSC1 scan_line 1,2,3,4
RSUS RSCAN
KSC2
KSC3 scan_line 3,4 KSC4
scan_line 4
CSC1
CSS
sus_line 2,3,4
scan_line 1,2,3,4
CC1 CS2
scan_line 2,3,4
scan_line 4
3D Field Simulation Model
CS1
CS3 sus_line 4
scan_line 2,3,4 CC2 scan_line 3,4 CC3
scan line1
CSC2 scan_line 2,3,4 CSC3 scan_line 3,4 sustain line 4 C SC4
scan_line 4 scan_line 4
scan line 4
Extracted Compact Circuit Model
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
90/97
RF PDP – Full Board Simulation of the RF PDP IV • DUT Model Electrical Parameter Extraction III – 28 Line Connector 28 line connector
KCNTC1
25 mm
LCNT1
RCNT1
CCNTC1 CCNT1
line 2,3,...,28 KCNTC2 line 3,4,...,28
line 1
line 2,3,...,28 CCNTC2 line 3,4,...,28
line 2
20 mm KCNTC27 line 28
CCNTC27 line 28 LCNT28
RCNT28
line 27
line 28 CCNT28
3D Field Simulation Model
Extracted Compact Circuit Model
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
91/97
RF PDP – Full Board Simulation of the RF PDP V • DUT Model Electrical Parameter Extraction IV – RF PDP Panel Test Frame
0.8 mm
12.5 mm 13.5 mm
3D Field Simulation Model
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
92/97
RF PDP – Experimental Verification I • Measured- and Simulated Results – Input Impedance of the RF PDP Panel 40 measured Z(real) simulated Z(real)
40
Impedance [Ω]
Impedance [Ω]
60
20
0
0 40 80
measured Z(imag.) simulated Z(imag.)
120 160
20
40 60 80 Frequency [MHz]
100
20
Real Part
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
40 60 80 Frequency [MHz]
100
Imaginary Part
Dec. 2004
93/97
RF PDP – Experimental Verification II • Display Image With and Without the Impedance Matching Block
Without I.M.B.
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
With I.M.B.
Dec. 2004
94/97
IV. Conclusions
Conclusions I • Full Board Simulation of the RF Systems – – – –
Based-on the Three Dimensional Field Analysis Segmentation Method Simulation Parameter Extraction Software-based De-embedding/ Impedance Analysis
– Optimization of the Test Fixture
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
96/97
Conclusions II • Designs and Verifications of Three Distinct RF Systems Using Proposed Full Board Simulation Method – CMOS RF IC • Electrically Small, But Complex RF System • Simulation Parameter Extraction for the Full Board
– RF LPA • Full Board Simulation of the LDMOS-based Power Amplifier • Propose Wide Dynamic Range Predistortion-type LPA
– RF PDP • Simple, But Repeated Large Size RF System • Simulation Parameter Extraction for the Full Structures
KiHyuk Kim, NanoElectronics Lab, Department of Electronics Engineering, Korea University
Dec. 2004
97/97