a new approach to low pin count products test found ... · vlct (texas instruments) ets-88 ......
TRANSCRIPT
A new approach to low pin count products test found Vertical Probes for
to be superior to Cantilever
Franz Steger (TI)Alessandro Antonioli (TP)
Raffaele Vallauri (TP)
Overview• Introduction• Methods and Materials• Results / Fields of comparison• Summary / Conclusion
2Franz Steger
Introduction• Texas Instruments• FTest = Freising Test (Freising, Germany)
3Franz Steger
Product portfolio:• High performance analog• Linear & Logic• Power management• …
*ADVANCES IN ELECTRONIC TESTING:CHALLENGES AND METHODOLOGIES, DIMITRIS GIZOPOULOS
Semiconductor markets, Moore’s Law, and testing*
FTest products
Methods and Materials• Introduction• Methods and Materials
– Test candidates– Probe solution– Technoprobe T1 needle– Technoprobe XLT option– ROI estimation
• Results • Summary / Conclusion
4Franz Steger
Methods and Materials• Test candidates
5
Single site Quad site 16 site
Product Op-Amp Mobile Application Little logic
Needles 61 96 80
Tester ETS-364(Eagle / Teradyne)
VLCT(Texas Instruments)
ETS-88(Eagle / Teradyne)
Max. Curr. 50mA 10mA 50mA
Max. Frequ. n.a. 3MHz 1GHz
Max. test temperature 125C 25C 85C
Franz Steger
Methods and Materials• Cantilever probe
Tungsten Rhenium
• FR4 probe card• Stiffener
(as requ. by vendor)
• Allied Diamond Lapping(50-30145 3um, PinkPad)
• Technoprobe T1 with XLT option
• FR4 probe card• Stiffener
(as requ. by vendor)
• MIPOX WA6000 SWE
6Franz Steger
Method and Materials• Technoprobe TPEG™ MEMS T1 needle technology:
– Fine pitch down to 55um– Low force for Al Pad probing and PoAA– Main characteristics:
7Franz Steger
PARAMETER TPEG™ MEMS T1
Needle diameter Less than 1,5 mils equivalent
Max pin count > 20.000 pins
X, Y alignment accuracy and Z planarity X,Y: ± 8 µm; Z plan: ∆ 20 µm
Min pitch and configuration
55 µm linear configuration
Pin Current (CCC) 410 mA
Force (at 3 mils OT) 2 g or 3 g
Method and Materials• Technoprobe XLT option
– Patented solution by Technoprobe– XLT option offers a longer usable tip while maintaining the
advantages of the already proven TPEG™ MEMS T1 characteristics
• Low and constant force• Effective cleaning recipe for stable CRES• No scrub probe marks
8Franz Steger
ROI estimation (4M+ TD)• Cantilever
– Initial cost X$– Renew Y$– Online cleaning Z$– Offline maintenance 1.5h
è Running cost K$
9Franz Steger
• TPEG™ MEMS T1 XLT– Initial cost 8*X$– Renew Y$*2– Online cleaning 0.67*Z$– Offline maintenance 0.3h
èRunning Cost ~0,1*K$
• TPEG™ MEMS T1 XLT Probe Card breakeaven cost vs cantilever at 2.000.000 touchdowns
Results• Introduction• Methods and Materials• Results / Fields of comparison:
– Probe marks– Contact resistance– Operating performance– Throughput– Cost of ownership
• Summary / Conclusion10Franz Steger
Results• Probe Marks Analysis
– TPEG™ MEMS T1 XLT technology outperforms cantilever in respect of probe marks and pad damage:
• No punch through and no cracks below bond pad found.
• 15 test runs on all test canditates
11
TPEG™ T1 XLT
Cantilever
Franz Steger
Results• Contact resistance (vertical only)
12
1st wafer 2nd wafer 3rd wafer
Cont
act r
esist
ance
(a
bsol
ute)
[Ω]
Franz Steger
• Cleaning recipe – Over travel (test): 80 um (first touch)– Online cleaning media: MIPOX WA6000 SWE– Cleaning freq.: 35 touchdowns– Strokes: 1 @RT, 2@ HT– Over travel (clean): 85 um
• è Contact resistance: Average 1.2 Ω , std. dev. 0.06 Ω (over 250k TDs)
• Offline Mantenance– Every 500.000 Touchdowns
• Visual inspection (incl. free length measurement)
• Particle removal (pressurized air, IPA, brush) if required.
[TDs]0 163130
163130
Results• Contact resistance distributions
– TPEG™ T1 XLT has much narrower distribution vs Cantilever enabling better control over time.
13
1st pin 3rd pin2nd pin
CantileverTPEG T1 XLT
CantileverTPEG T1 XLT
CantileverTPEG T1 XLT
[Ω] [Ω] [Ω]
Franz Steger
Results• Operating performance
14
0,00
2,00
4,00
6,00
8,00
10,00
12,00
1-Ap
r
2-Ap
r
3-Ap
r
4-Ap
r
5-Ap
r
6-Ap
r
7-Ap
r
8-Ap
r
9-Ap
r
10-A
pr
11-A
pr
12-A
pr
13-A
pr
14-A
pr
15-A
pr
16-A
pr
17-A
pr
18-A
pr
19-A
pr
20-A
pr
21-A
pr
22-A
pr
23-A
pr
24-A
pr
25-A
pr
26-A
pr
27-A
pr
28-A
pr
29-A
pr
Number of unplanned interruptions per card under test
Canti
Vertical
Franz Steger
Start End
Results• Impact on throughput
15
0
2
4
6
8
10
12
14
Average troughput per tester
Canti
Vertical
Franz Steger
Start End
Results• Lifetime
16
4
6
8
10
12
14
16
0 500.000 1.000.000 1.500.000 2.000.000 2.500.000 3.000.000 3.500.000 4.000.000 4.500.000 5.000.000
Rem
aini
ng n
eedl
e le
ngh
[mil]
[Touchdowns]
#1
#2
#3
#5
#4
EOL
Franz Steger
ROI estimation (4M+ TD)• Cantilever
– Initial cost X$– Renew Y$– Online cleaning Z$– Offline maintenance 1.5h
è Running cost K$
17Franz Steger
• TPEG™ MEMS T1 XLT– Initial cost 8*X$– Renew Y*0.16$– Online cleaning 0.67*Z$– Offline maintenance 0.3h
èRunning Cost ~0,1*K$
• TPEG™ MEMS T1 XLT Probe Card breakeaven cost vs cantilever at 2.000.000 touchdowns
Results• Financial Impact
18Franz Steger
Conditions for calculation- Cantilever cards already in house (no new acquisition, only rebuilt / refurbish at EOL)- PCB is re-used for vertical
0,00
5,0010,00
15,00
20,0025,00
30,0035,00
40,0045,00
50,00
0 1 2 3 4 5 6
Cost
s [k$
]
[Touchdowns]Milioni
Total Costs for Probe
TPEG T1 XLT
Cantilever
SummaryConclusion:• High performance probe solutions are cost effective also for low
pin count probe card applications• This study has demonstrated the advantage of using vertical
technology (Technoprobe T1 XLT) also for low pin count probe cards:
– No pad damage– Stable contact resistance with better distribution– Higher life time with less maintenance effort– Higher initial cost but overall lower cost over lifetime à Better ROI
19Franz Steger
Acknowledgement• Texas Instruments
– Dirk Jasmer, Maciej Miler (Product responsible)– Al Wegleitner (& PTS team)– Werner Huber (& FTest management team)
• Technoprobe– Detlev Koch (Teltec)– Alessandro Antonioli– Raffaele Vallauri
20Franz Steger
Thank you for your attention
21Franz Steger