ADS Simulation Experience in Mobile Phone PCB … Chung Outline • About FIH & me • RF matching...
Transcript of ADS Simulation Experience in Mobile Phone PCB … Chung Outline • About FIH & me • RF matching...
ADS Simulation Experience in Mobile Phone PCB
MatchingTommy Chung
Outline• About FIH & me• RF matching and tuning• Other people simulation process
• Case1• Case2
• What I think about simulation• Disassemble simulation schematic• My Simulation step
• Prepare• Process • compare• Set up
• Case and key point• Conclusion• Summary
About FIH & me
• Foxconn international holding(富智康)• 新北市土城區民生街四號(靠近頂埔捷運站)• Main :
• Smart phone, wearable device.• Me
• Now• RF engineer
• Past• Power semiconductor(SiC)• photonic crystal• Antenna simulation
RF matching and tuning
Other simulation process: case 10.
5
1.0
2.0
5.0 10 20
20
-20
10
-10
5.0
-5.0
2.0
-2.0
1.0
-1.0
0.5
-0.5
0.2
-0.2
freq (1.700GHz to 2.000GHz)
A1M
_PC
S_D
CS
_PA
OU
T_a
dfi..
S(9
,9)
Other simulation process: case 2Start
Use RF cable and NA to extrction trace loss and S-
parameter
Combine trace and parts S2P to get Schematic
End
Matching trend
Outline• About FIH & me• RF matching and tuning• Other people simulation process
• Case1• Case2
• What I think about simulation• Disassemble simulation schematic• My Simulation step
• Prepare• Process • compare• Set up
• Case and key point• Conclusion• Summary
What I think about simulation
• Less welding • Reduce damage
• Momentum • For ours simulation is reliable or not.
• Component S-parameter • Should I measure it?
• Layout stack up • How to describe to ADS?
Disassemble simulation schematic
50ohm
Z = 50ohm
Zant= 50W || 7.5 nH
Ztx = 50W + 4.0 nH
Zrx = 50W
Z = 50ohm
conjugate
0.5
1.0
2.0
5.0 10 20
20
-20
10
-10
5.0
-5.0
2.0
-2.0
1.0
-1.0
0.5
-0.5
0.2
-0.2
freq (500.0MHz to 2.000GHz)
conj
(S(1
,1))
S(1
,1)
Load pull
chip
wire
C and L
My Simulation-prepare
1.51 1.52 1.53 1.54 1.55 1.56 1.57 1.58 1.591.50 1.60
-0.087
-0.086
-0.085
-0.084
-0.083
-0.082
-0.088
-0.081
freq, GHz
dB(A
1M_G
PS
_AN
T_a
dfi..
S(1
,2))
0.5
1.0
2.0
5.0 10 20
20
-20
10
-10
5.0
-5.0
2.0
-2.0
1.0
-1.0
0.5
-0.5
0.2
-0.2
freq (1.500GHz to 1.600GHz)
A1M
_GP
S_A
NT
_adf
i..S
(1,1
)A
1M_G
PS
_AN
T_a
dfi..
S(2
,2)
My simulation-process
Start
Extraction Adfi file,and load adfi to ADS
Load adfi file to ADS
Set up simulation parameter
Run momentum
load adfi to A S
Load adfi file to ADS
End
Set up simulation parameter
Run momentum
Place component on schematic
Get insertion loss and smith chart
My simulation-compare
Start
Use RF cable and NA to extrction trace loss and S-
parameter
Combine trace and parts S2P to get Schematic
End
Matching trend
Start
Extraction Adfi file,and load adfi to ADS
Load adfi file to ADS
End
Set up simulation parameter
Run momentum
Place component on schematic
Get insertion loss and smith chart
My simulation-set up
• Mesh not over 60000 matrix• Stack must follow real stack• Less stack up is best• Port and GND follow real• Frequency not to wide• Thick:2D,via:lumped,mesh:10• Please use Queue and auto create EM model• Remember place real component to your schematic
Outline• About FIH & me• RF matching and tuning• Other people simulation process
• Case1• Case2
• What I think about simulation• Disassemble simulation schematic• My Simulation step
• Prepare• Process• compare • Set up
• Case and key point• Conclusion• Summary
Case and key point-time
Simulation time compare(mesh >60000) mesh 30000CPU Intel Xeon @2.6GHz(8 core)*2 8 core i5(4 core) i5(4 core)
RAM RAM: 132GB 32GB 16GB 16GB
speed time 1hour 8hours >12 hours 0.5 hours
Case and key point-thick,via
• Physical Model• Keysight thick : 2D ,via : 2D• Me thick : 2d , via : lumped
Case and key point-thick,via
Smith chart compare
Case and key point-thick,via
Insertion loss compare
Case and key point-component
• Conponent : ideal or real
Ref
1 2
Term1 Term3
SNP1
Term5 Term7C1
Num=1
Z=50 Ohm
Num=3
Z=50 Ohm
File="D:\conponent_package\0.5pF.s2p"
Num=5
Z=50 Ohm
Num=7
Z=50 OhmC=0.5 pF
Term Term
S2P
Term TermC
0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.80.6 3.0
-12
-10
-8
-6
-4
-14
-2
freq, GHz
dB(S
(1,3
))dB
(S(5
,7))
0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.80.6 3.0
-0.06
-0.04
-0.02
-0.08
0.00
freq, GHz
dB(S
(2,4
))dB
(S(6
,8))
0.5pF
100pF
Ref
1 2
Term1 Term3
SNP1
Term5 Term7L1
Num=1
Z=50 Ohm
Num=3
Z=50 Ohm
File="D:\conponent_package\HKQ0603U_0N6.s2p"
Num=5
Z=50 Ohm
Num=7
Z=50 OhmL=0.6 nHR=
Term Term
S2P
Term TermL
0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.80.6 3.0
-40
-35
-30
-25
-20
-15
-10
-45
-5
freq, GHz
dB(S
(2,4
))dB
(S(6
,8))
0.6nH
39nH
0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.80.6 3.0
-0.05
-0.04
-0.03
-0.02
-0.01
-0.06
0.00
freq, GHz
dB(S
(1,3
))dB
(S(5
,7))
Case and key point-gnd
Case and key point-Stack up
prediction
finalFinal in ADS
Case and key point-Stack up
Case and key point-Stack up
Case and key point-trace loss
• The trace loss is about 0.2dB(2GHz)
Case and key point-trace loss
• Layer2 GND
• Add two via from • layer2 to layer3
• Add two via from • layer2 to layer1
Conclusion
• The skill is enough to estimate the insertion loss and smith chart for ours matching.
• If we want to get total trace loss or smith chart accurately, we should measure the s-parameter of chip.
• Anyone want to simulate more region ,maybe use server or SIPRO.
Summary
• Mesh• Real Stack up• GND port• Frequency • Thick:2D,via:lumped,mesh:10• Queue• real component
Q&A