feinprobe® solution for wlcsp applications - swtest.org · the standard procedure of international...
TRANSCRIPT
FeinProbe® Solution for WLCSP Applications
Krzysztof Dabrowiecki
2
Overview• Motivation• WLCSP Market Drivers• WLCSP Application Requirements• Stages of FeinProbe® Internal Qualification and Testing• Summary and Conclusion
K. Dabrowiecki
Motivation• Diversification of semiconductor products:
– Smartphones/Tablets– Internet of Things/Wearables– Virtual Reality– Artificial Intelligence– 5G Connectivity– Automotive and much more others like Medical, Industrial, Military…
• Increasing WLP (Fan-in and Fan-out) without substrates or interposer for IC chips
• Increasing final tests at wafer level
3K. Dabrowiecki
WLCSP Market Drivers• Good electrical performance • Good thermal performance• Smaller and lower profile• High power• High system integration• High I/O density• Low cost
4K. Dabrowiecki
General WLCSP Requirements
Low resistance for DC measurements High bandwidth for functional tests or AC
tests Low path inductance to reduce the noise
on the DUT power and ground at its switching frequency
High current capacity for power delivery and DC parametric tests
Kelvin probes for analog and mixed signal
5K. Dabrowiecki
Electrical Mechanical High contact force Short probe length for electrical
parameters Large plunger compliance for
reliable contacts Longer lifetime Effective cleaning method
Die Size< 100 sq mm
Low I/O Count < 500 Pin Count
Bump Pitch> = 300 µm
No IC Underfill and no IC Substrate or Interposer
6K. Dabrowiecki
Fan-out WLP Connection to PCB
IC Package Mold
Board
RDL Thin Film
Probe Card Technology for WLCSP Testing
7K. Dabrowiecki
Contact Force Direction
Electrical Contact
Contact Force Direction
Contact Force Direction
Electrical Contacts
Electrical Contact
Cantilever Technology Buckling Beam/Hybrid BB Technology
Spring Probe Technology
FeinProbe® Spring Contact Advantage
8K. Dabrowiecki
Stable and consistent contact resistance Self-aligning plunger tip High bandwidth High Current Carrying Capability (CCC) Large spring range force Large plunger tip compliant Low bump damage No probe coating required
Crown Plungers of FeinProbe® Head
9K. Dabrowiecki
Stages of FeinProbe® Qualification and Testing
10K. Dabrowiecki
Finite Element of Solder Bump-Crown Tip Plunger Model
Contact Force Test of Probe Head
Path Resistance on Bumped Wafer
Vector Network Analyzer Probe Test
Probe Current Carrying Capability Method
Why Finite Element Bump-Probe Contact Study?
11K. Dabrowiecki
CSP with RDL:- 4 mask- 9 um UBM Cu/Ti
Low cost solution: - No RDL- 2 mask- 5 um UBM Au/Ni/Cu/Ti
Current WLCSP Technology Options
Source: C.Lee Amkor
Finite Element Probe-Bump Models
12K. Dabrowiecki
FE model – Quarter
Lead-free bump diameter 142um
Crown plunger diameter 160um
Plunger force F=20gf
FE Model w/ 2 maskFE Model w/ 4 mask Boundary Conditions
Model Material Properties Table
13K. Dabrowiecki
Model part MaterialYoung's
Modulus E (GPa)
Poisson's Ratio (-)
Yield Point (MPa)
CTE (ppm/C)
Crown plunger Pd alloy 110 0.35 930 11.7
Solder ball Sn/Ag 50 0.36 34.2 25Cu Cu 117 0.34 62 16.8PBO 1 PI 2.3 0.3 69 35PBO 2 PI 2.3 0.3 69 35Custom Passivation PI 3.5 0.35 69 35Silicon Si 131 0.26 UTS 7000 2.54Gold Au 79 0.4 205 14.2Nickel Ni 207 0.31 138 13.3Titanium Ti 116 0.32 379 8.5
Source: Amkor, NIST, STATS ChipPAC, Japan Institute of Metals
Model Stress Simulation
14K. Dabrowiecki
FE Model w/ 2 maskFE Model w/ 4 mask
Stress Calculation in the UBM
15K. Dabrowiecki
FE Model w/ 2 maskFE Model w/ 4 mask
High stresses at the bottom of UBM and on the edge
High stresses underneath of solder bump
Potential Bump Failures
16K. Dabrowiecki
Crack from UBM edge UBM and RDL detachment
Source: Journal of Electronic Packaging, 2015
Silicon die
The failure causes are related to high range of temperature and possible high stress concentrations in the UBM
0
2
4
6
8
10
12
14
16
18
0 10 20 30 40 50 60 70 80 90 100
110
120
130
140
150
160
170
180
190
200
Cont
act F
orce
[gf]
Over Drive [um]
Contact Force Test - Full Probe Head
17K. Dabrowiecki
The contact force of test probe head measured by EAP500 system at RT
Individual pre-loaded spring probe
Average probe contact force installed in probe head
Plunger friction in the probe head
Path Resistance Test
18K. Dabrowiecki
The bumped wafer SAC305: pitch 250um, bump size 142umThe test over drive 175um
Path Resistance Illustration
19K. Dabrowiecki
Shorted Wafer
Probe Head
J750 Wired Connector
Signal PathReturn Path
Prober Head
The path contact resistance contains material resistivity of wafer, solder bump, probe, connector wire, PCB solder, PCB trace and prober head pogo, and contact resistances of probe to bump, probe to wired connector, pogo to PCB and pogo to probe head
Path Resistance and Life Test Charts
20K. Dabrowiecki
At Varied Temperature and OD Life Test at 175um OD
Path resistance at 125C is slightly higher (0.2 Ohm) than at RT and -40C
Path resistance at 175um OD shows a quite uniform and consistent profile up to 500k TD’s
Wafer Bump Image at RT
50µmOver Drive 100µm 150µm 175µm
Wafer Bump Image at 125C
Wafer Bump Image at -40C
Wafer Bump Scrub Marks
21K. Dabrowiecki
Vector Network Analyzer Test
22K. Dabrowiecki
Test fixture with VNA cables
Configuration: G-S-GProbe pitch: 400um
ENA 5071C
SMA JACK, 50 Ohm, 27GHz
SMA JACK, 50 Ohm, 27GHz
Test FixtureGG S
Impedance and s-Parameters Measurements
23K. Dabrowiecki
The impedance characteristic was obtained by applying a signal with rise time 50ps (20-80%)
S21- Insertion Loss: 9.3GHz @ -1dBS11 – Return Loss: 9.5 GHz @-10dB
Probe Current Carrying Capability - Method
24K. Dabrowiecki
The standard procedure of International Sematech Manufacturing Initiative (ISMI 2009) describing the probe current carrying capability test methods has not mentioned anything about a method how to define the CCC for spring probe.
Therefore based on our internal empirical researches and tests, the maximum continuous current has been defined as a maximum tolerable probe warming temperature during the current lab test.In case of FeinProbe®, the temperature change of probes under the current has been defined for 20 degree (K) measured by thermocouples
Tp = K + TeTp – probe surface temperatureK – probe temperature change measured by thermocouples Te – environment temperature (RT)
Probe Current Measurement Cases
25K. Dabrowiecki
1 - power supply 2 – digital voltmeter3 – thermocouple meter
Thermocouples touching the test probe
Thermocouples with distance to the test probe
4 - thermocouples 5 - spring probe6 - probe station
Schematic Diagram Description
Case 1 Case 2
0
10
20
30
40
50
60
70
80
0.2 0.3 0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 2.1 2.3 2.5 2.6 2.9 3.1 3.3 3.5 3.7
Tem
pera
ture
Cha
nge
[K]
Current [A]
TC with Distance TC Touch
Probe Temperature Change Chart
26K. Dabrowiecki
+20%
Probe Image in the Eye of Infrared Camera
27K. Dabrowiecki
Probe current I=1.7A
Probe Surface Temp Measured by IC
28K. Dabrowiecki
3.7A
3.4A
3.2A
2.6-2.8 A
2.2A
0.4 – 1.9ANo color change
No color change
Slightly color change
Color changes along entire barrel
Black color
Red color
Note: Probe surface observation during the current test with TC. Test time for each current cycle t=2 minutes
0102030405060708090
100110120130140150160170180
0 0.5 1 1.5 2 2.5 3 3.5 4
Prob
e Su
rfac
e Te
mpe
ratu
re [C
]
Current [A]
FeinProbe® Current CC and Resistance Graphs
29K. Dabrowiecki
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0
20
40
60
80
100
120
140
0 1 2 3 4 5 6
Path
Res
ista
nce
[Ohm
]
Tem
pera
ture
cha
nge
[K]
Current [A]
Temperature change Resistance
Probe length L= 8.6mmContinuous CCC= 2.1 APath resistance = 70 mOhm
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0
20
40
60
80
100
120
0 1 2 3 4 5 6 7
Path
Res
ista
nce
[Ohm
]
Tem
pera
ture
cha
nge
[K]
Current [A]
Temperature change Resistance
Probe length L=3.2mmContinuous CCC= 2.8 APath resistance = 65 mOhm
Lowest Probe Resistance Area Lowest Probe Resistance Area
Summary and Conclusions
30K. Dabrowiecki
- The advanced package technology exploring the 3D, complex and
high pin count of FOCSP requires a reliable test solution and the
presented five stages of qualification methodology can be a good
approach to guarantee an appropriate contact solution
- The thinner and smaller UBM may cause a cracking or
delamination issues of the under bump metallization during the
wafer test
31K. Dabrowiecki
Summary and Conclusions
- The update of ISMI procedure is needed to standardize the CCC
measurement method for the spring probes. The presented
approach is a suggestion for an alternative method
32K. Dabrowiecki
Follow-on Work
- Study of the low spring force probe
- Study of the spring probe at high temperature
Acknowledgments
33K. Dabrowiecki
Thank You
I would like to thank colleagues at Feinmetall for their help in
preparation of this presentation. Many thanks to Jörg Behr,
Gunnar Volkmann and Jörg Burgold.