Best Practices Factory

Implementing Pb-Free
Assembly at Your Factory

Ronald C. Lasky, Ph.D., PE
Senior Technologist Indium Corp
Visiting Professor, Dartmouth
Timothy Jensen
Indium Corp
Feb 2004
Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Agenda
• Pb-Free World Status
• Alloy Selection
• Getting Your Facility Ready: Best Practices
• PWB and Component Finishes
• Solder Paste Pre-Screening
• Overview of Motorola Implementation
• Large Board and Wave Soldering Issues
All slides available at www.pb-free.com
Slide #1 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Dr. Ron Lasky
• Native of Binghamton
• Graduate of BCC, Cornell,
BU, Cornell
– PhD in materials science
• NYS Professional Engineer
• More than 20 years in
electronic and
optoelectronic packaging
at IBM, Universal
Instruments, Cookson
(Alpha)
• Author of 5 books
• Currently a Senior
Technologist for Indium
and a Visiting Professor at
Dartmouth College

Electronics:
Still the Largest Industry
• Holding at $1 trillion
– Bigger than automotive
– But Food?
• $170/yr x 6 billion > $1 T
• Typical long term growth is 6-8%
CAGR
– Vs 2-4 % for economies in good times
– Early 2000s down slightly

The Electronic Industry: $1.1T
Military, $89
Automotive, $52
Industrial and Medical,
Computer, $429 $113
Consumer, $123
Communications, $300
All assembled on 30-35K lines!

Source: Prismark

Courtesy Prismark

System Cost: $1K PC
Discretes: 4%
ICs: 27%
Substrates: 4%
Assembly & Test: 14%

SGA & Profit: 16%
Connectors: 3%:
Distribution: 14%
R&D: 2% Solder: 0.05%!
Housing and I/O: 16%
Source: Prismark

What is a Printed Circuit Board (PCB)?
• Starts with a PWB • Reflow melts the solder
• Solder Paste is printed • The finished PCB is
• IC, passive, active tested
components are placed • Thru hole assembly may
also be performed
Source: The Internet

Cross Section of a PQFP Component
Molding Compound Wire Bonded
Silicon Die
Nitto Denko Tanaka
8 cents 18 cents
Die Attach Adhesive Courtesy:

Ablestik Prismark
0.4 cents
Leadframe
Shinko 1.2 cents
91 cents per lead
The silicon die is the heart of electronics, it produces all of

digital and analog functions. The material and leads that
enclose the die are called the package. The resulting device
called a component, a component package or a semiconductor
package.

• The relentless size
reduction of
passives
• Assembling the
0201 is one of the
current process
challenges
• A trillion
assembled each
year Source: The Internet

PWB Assembly
Stencil Chip Shooter Pick and

Reflow Oven
Printer Place
Hand Mount and Opto

Wave or Selective Inspection/Test
Sources: DEK, UIC, Electrovert Solder

Pb-free Initiatives & Global
Trends

Current Pb-free Legislation
• WEEE/ROHS directives
• Japanese Home Electronics

Recycling Law
• USA? Some state activity, nothing

federal
– EPA: ‘Deal with Electronics or We Will’

Pb-free Legislation
• EU Ratifies WEEE / ROHS directive
– Dec 2002 EU parliament passed legislation
to ban use of:
• Pb, Hg, Cd, Cr VI, PBB, PBDE
– Affects ALL products sold after July 2006.
• Exceptions
– Telecom equipment until 2010
– High lead solder applications (>85% lead alloy)
– Recycling program enforced June 2005.

Pb-free Legislation
• Japanese Pb-free Activity
– No direct ban on using Pb
– MITI proposed recycling legislation in May 1998.
– Japan home electronics recycling law requires
OEMs to collect and recycle 4 major products
since April 2001.
– These measures, as well as marketing advantage,
are pushing major Japanese companies to be
more environmentally conscience

Japanese Pb-free Products
• Matsushita / Panasonic
announcement
– By March 2003 – Over 70 million PCBs
(approximately 12,000 product models) will
be built using a Pb-free solder.
• Products have been built in over 100 different
factories

Panasonic Pb-Free Examples
MJ70 Minidisc Player MJ30 Minidisc Player

Hitachi Pb-Free Examples
H845L Camcorder PC Audio Board, Flora 220CX

Is Pb-Free a good thing?

True or False
• Electronics is a major contributor
to environmental lead. ⇒
FALSE
– Batteries: 4,000,000 tons

– Bullets: 200,000 tons
– Electronics: 18,500 tons < 0.5%
– Not to mention tire balancing weights!
• About 100,000 tons/year

True or False
• The level of lead in the blood of
US residents has increased
since 1945 ⇒ FALSE
• Paint and gasoline reductions

have had a stunningly positive
effect.

Average US Lead Levels in Micrograms per
Deciliter of Blood
30
25
20
Pb( µg/dl)
15
10
0
1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995
Year

Pb does Leach
• Pb Leach Testing
– Use water with a pH of 7-7.05
– Pb will not leach very much at this pH
• In acid water, Pb does leach
• After all, we banned Pb from
plumbing for good reason
• However, no measurable results in a
“real life” experiment

Pb-Free Solder Alloys
• Present alloys chosen to replace
Sn/Pb are Sn/Ag/Cu alloys
• Ag-containing Solders also Leach
– EPA Groundwater Leaching Tests
• ALL Pb-free silver-containing solders fail
– “Lead-free Solders: A Push in the Wrong Direction?”
Ed Smith, K-Tec Inc.
– “Reliability & Leachate Testing of Pb-free Solder
Joints” – Thomas Woodrow, Boeing

Pb-Free Solder Alloys
• Silver containing replacements
– Are we taking a wrong turn?
• Silver is toxic
• Silver is a biocide
• Silver is known to kill more than 650 different viral,
bacterial and fungal organisms
• Also, Silver will kill most marine life

Pb-Free Initiative Summary
• Simply changing from SnPb to SnAgCu may not fix
the problem
• Recycling must be the long term answer
• Expect more legislation and higher landfill costs in
the years to come!
Despite the data, Pb-free solder in the Electronics

Industry is going to happen…

Pb-Free Alloy Selection

Alloy Selection Overview
• Potential Alloy Overview
• General Alloy Search
• Sn/Ag/Cu in Depth

Alloy Decision Process
• The melting point should be as close to
Sn/Pb eutectic as possible
• The alloy must be eutectic or very close to
eutectic
• It must contain no more than three
elements
• Use of existing patents should be avoided
to ease implementation
• Reliability should be equal or better than
Sn/Pb Edwin Bradley – NEMI chairperson, Motorola

Options
• Current Alloy List
• Alloy properties
• Pros & Cons of each

Options: Current Alloy List
– Sn/Ag – Sn/Ag/Bi/X
– Sn/Cu – Sn/Sb
– Sn/Ag/Cu – Sn/Zn
– Sn/Ag/Cu/Sb – Sn/Bi

Options: Current Alloy List
– Sn/Ag – Sn/Ag/Bi/X
– Sn/Cu – Sn/Sb
– Sn/Ag/Cu – Sn/Zn
– Sn/Ag/Cu/Sb – Sn/Bi

Scientific Realities: 91Sn/9Zn
• Melting point = 199°C
• Zinc reacts with both acids and bases
• All flux/vehicles are mildly acidic at room
temperature
• Which means shelf-life is in the order of
days, not months!
• Also may need conformal coating once in
place - Zn oxidizes very readily

Scientific Realities: 96.5Sn/3.5Ag
• Melting point = 221°C
• Years of experience
• Relatively poor wetting
• Poor thermal cycling (-40/+125)
1st Last
Alloy # on #
Composition Failure Failure Rank
Code Test Failed
(cycle) (cycle)
A1 Sn 3.5Ag 12 12 1282 1987 8
A11 Sn 4Ag 1Cu 14 14 2340 2552 3
A14 Sn 4Ag 0.5Cu 14 14 2108 2579 5
A21 Sn 2.5Ag 0.8Cu 0.5Sb 14 14 2378 2378 2
A32 Sn 4.6Ag 1.6Cu 1Sb 1Bi 15 15 2161 2161 4
A62 Sn 3.4Ag 1Cu 3.3Bi 14 14 1864 2527 6
A66 Sn 3.5Ag 1.5In 14 14 2387 2577 1
B63 Sn/Pb Control 13 13 1845 2607 7 Source: NCMS

Scientific Realities: 58Sn/42Bi
• Melting point = 138 °C
• Years of experience
• Poor shock resistance
• 1% Ag addition strengthens this alloy
• Low MP eliminates temperature issues
with components and board

Scientific Realities: Sn/Ag/Cu
• Melting Point = 217°C
• Cu improves wettability, creep, thermal
fatigue
• High solderability & reliability

Scientific Realities: Sn/Ag/Cu/Sb
• Melting point = 215 – 217°C
• 0.5% Sb may strengthen alloy
• May be considered for wave soldering by
effectively lowering Ag content
• Patented composition may limit worldwide
availability
• Four part alloy makes manufacturing
consistency less reliable.

Scientific Realities: Sn/Ag/Bi
• Melting Point = 210 - 215 °C
• Bi lowers melting point & improves
wettability of SnAg alloys
• Avoid using in presence of Pb:
• a 96°C ternary alloy(Sn/Pb/Bi) may form

Pb-free Alloy Investigation
• Phase 1: DOE to analyze 10 probable

alloy solutions using Sn/Pb as
Benchmark
• Phase 2: Optimize flux chemistry

Alloy Performance Investigation
Alloy Solidus Liquidus
63Sn37Pb 182.1 183.0

96.5Sn3.5Ag 219.7 220.8
99.3Sn0.7Cu 225.7 227.0
95.5Sn3.8Ag0.7Cu 216.3 217.5
93.6Sn4.7Ag1.7Cu 215.9 217.1
96.2Sn2.5Ag0.8Cu0.5Sb 216.9 218.2
91.7Sn3.5Ag4.8Bi 202.1 215.1
90.5Sn7.5Bi2Ag 190.6 214.7
58Bi42Sn 136.3 138.5
95Sn5Sb 238.3 240.3
89Sn8Zn3Bi 190.6 195.4

Pb-free Alloy Investigation
• 5 Individual studies performed

– Wetting
– Shelf life
– Tack time
– Solder Ball
– Visual reflow inspection
• Performance ranked 0-10 for each
category

R&D Test #1: Wetting
Cu pad Unwetted Cu
70%
Print Reflow spread

paste
(offset) Solder
bump

Wetting Examples
4 5 7
40% 50% 70%

R&D Test #2: Shelf Life
Shelf life was based on the percentage change
in viscosity over time. The length of the test
was 30 days.
SLI Description
0 Overall instability > 25%
2 Overall instability = 20-25%

R&D Test #3: Tack Stability
6
8
10
4
Tack 2
0
0
0 Time
R&D Test #4: Solder Balling
SBI Number of solder balls
0 Did not reflow
1 > 501, with some reflow
2 401-500
3 301-400
4 201-300
5 151-200
6 101-150
7 51-100
8 21-50
9 11-20
10 0-10

Alloy Investigation- Results
1. Sn63/Pb37 7. Sn99.3/Cu0.7
2. Sn91.7/Bi4.8/Ag3.5 8. Sn96.2/Ag2.5/Cu0.8/Sb0.5
3. Sn90.5/Bi7.5/Ag2 9. Sn95/Sb5
4. Sn95.5/Ag3.8/Cu0.7 10. Sn96.5/Ag3.5
5. Sn42/Bi58 11. Sn89/Zn8/Bi3
6. Sn93.6/Ag4.7/Cu1.7

R&D Test #5: Visual
BEST
10
2 BAD

Winner- Sn/Ag/Bi ???
• Bi addition lowers melting
point & improves wetting
• BUT……..Concerns with
• Fillet Lift
• Low temperature ternary alloy
formation with Pb contamination

Pb Contamination Concerns
Why Bi alloys are not the short-term solution
250
0% Pb
225 3% Pb
6% Pb
Onset Temp, deg. C
200
175
150
125 Sn-Pb-Ag
100 Eutectic
Sn-Bi-Pb
75
Peritectic
Sn-Bi-Pb Eutectic
50
Sn-3.8Ag- Sn-2Ag-2Bi Sn-2Ag-4Bi Sn-2Ag-7.5Bi Sn-10.5Bi Sn-12Bi
0.7Cu

Sn/Ag/Bi on Sn/Pb Finish
Sn/Ag/Bi with Pb contamination Sn/Ag/Bi with Pb contamination

after 1 reflow after thermal cycling
Ref: Zequn Mei, Fay Hua, and Judy Glazer, “SN- BI- XSOLDERS”, SMTA International, San Jose, CA, Sept. 13- 17, 1999.

Sn/Ag/Bi Fillet Lift
Ref: “Lead-Free Solder Project Final Report”, NCMS Report 0401RE96 , August 1997.

Bismuth Alloy Summary
• Sn-Ag-Bi alloys show formation of:

– Eutectic 96C phase at Bi > 10.5%Bi
– Peritectic 135C phase Bi > 4%Bi
In the presence of Pb.
• Sn-Ag-Cu alloys show formation of:
– Eutectic 179C phase.
• Wetting occurs at liquidus, so it is key melt
temp variable.
• PCB finish can affect wetting temperature due
to base metal diffusion into solder.

Conclusion
Pb-free Paste - SnAgCu
• Challenges:
– higher processing temperatures
– wettability
– must wet to a variety of metallizations

NEMI Alloy Summary
• NEMI Chosen Alloy
– 95.5Sn/3.9Ag/0.6Cu
– 217 C liquidus
• Pb-Free is now driven by both market factors
and now legislation
• SnAgCu preferred short/medium term solution
– NEMI, JEITA, IDEALS all agree on Sn/Ag/Cu
• When components become completely Pb-Free,
SnAgBi may become preferred solution
– Lower process temp
– Excellent wettability

Eutectic of Sn-Ag-Cu solders
B. Boettinger, K. W. Moon of NIST performed studies to determine the true Sn-
Ag-Cu eutectic.
Estimation of Ternary Liquidus Surface, 10/23/99
Based on Marquette saturation data,
with NWU and NIST thermal analysis.
8
270 C
7
6
250 C Ag3Sn
5
wt%Ag
230 C
4
220
3 C Cu6Sn5
230 C
2
Sn
Alloys in shaded 1
NIST experimental work
270 C
250 C
23 0
C showed that the composition
290
area have freezing

310
C
0 is approximately
range <10°C.
0 0.5 1 1.5 2 2.5 Sn3.5Ag 0.9Cu. (+/- 0.1%)
wt% Cu (In agreement with Loomis and Fine)

Relative Material Costs
• Sn $5.06/kg
• Zn $1.06/kg
• Cu $1.94/kg
• Bi $5.63/kg
• In $250.00/kg
• Ag $180.00/kg
• Sb $1.75/kg
• Pb <$1.00/kg

Pb-Free Cost
• All the lead-free alloys are more
expensive than the lead-containing
alloys
– Raw material costs are higher
– Currently there is no economy of scale
– Minimal Production experience
compared to Sn/Pb powder production

Which SAC to Use?
• Melting Point Comparison
• Wetting Comparison
• Long Term Cost Comparison

DSC of Actual Samples from
Solder Vendor A
96.5Sn/3.0Ag/ 95.5Sn/4.0Ag/ 95.5Sn/3.8Ag/
0.5Cu 0.5Cu 0.7Cu
Sample 1 216.76 °C 216.89 °C 216.70 °C
Sample 2 216. 49 °C 216.38 °C 216.56 °C
Sample 3 216.71 °C 216.35 °C 216.75 °C
Average 216.65 °C 216.54 °C 216.67 °C

IPC DSC Comparing SAC305
Suppliers
Supplier 1 Supplier 2 Supplier 3
Sample 1: 226.6 °C 217.8 °C 216.0 °C
Sample 2: 220.3 °C 216.7 °C 216.0 °C
Average: 223.45 °C 217.25 °C 216.0 °C

Source: IPC-SPVC-WP-006

Comparing the Wetting Forces
Difference in wetting per J-STD test procedure not

statistically significant between SAC alloys.
Source: IPC-SPVC-WP-006

SAC Actual Wetting
Source: CEMCEX2003 Seelig et al

Long Term Relative Costs
• Sn/Pb = 1x
• SAC305 = 2.15x
• 95.5Sn/3.8Ag/0.7Cu = 2.3x
• SAC405 = 2.35x

Alloy Summary
• Sn/Ag/Cu (SAC) is the best current option
• Sn/Ag/Bi worth consideration is Pb
contamination is not an issue.
• Sn/Bi doped with 1% Ag could be a viable
option for assemblies with tight
temperature restrictions.
• SAC305 ranks slightly higher than other
SAC alloys due to cost and performance.

Getting Your Facility Ready:
Pb-Free Implementation
Best Practices

Pb-Free Best Practices
• Assess Yourself
– Crucial with Pb-free
• Throughput Maximization
• CIP
• Process Software to Help
• Use the Right Tools
– Statistical Thinking
– DOE
– SPC

An SMT Process
Equipment
A
Product Product
Requirements
People Activities Procedures
Meets
Requirements
•On time
Materials
•Under cost

Assessment Categories
• DFM, Process and Equipment
• Materials Supply and Validation
• DOE, SPC, CIP
• Training and Failure Analysis
• Developed from pooled information
from industry experts

DFM, Process and Equipment
Ranking Key: 0 = strongly disagree or don't know , 3 = disagree, 5= neutral, 7 = agree, 10 = strongly agree
1 We have a documented and functioning DFM system that includes design ground rules (DGRs). Our entire 10
organization is strongly committed to DFM. DFM is a way of life (this question counts 3 times):
2 Our DGRs are established by using designed experiments and statistical process control: 8
3 The process engineers know how to run all of the equipment: 3
4 The engineer(s) responsible for stencil printing knows how to design a stencil 9
5 The engineer(s) responsible for component placement knows how to balance and optimize the placement 9
equipment. He assures that this operation is performed on all jobs:
6 The engineer in charge of the reflow process assures that the reflow profile matches the solder paste 9
specification:
7 Our process engineers have a disciplined and proven strategy to improve productivity: 9
8 Our process engineers have a disciplined and proven strategy to improve quality: 9
9 Our process equipment is "qualified" with a test and evaluation procedure that is founded on DOE principles: 3
10 There is a process engineer or team of engineers responsible for implementing new processes and technology: 9
Total Score out of 120 98
Ratings:
World Class = > 95 Your score places you as "World Class" in DFM, Process and Equipment for
Above Average = 75 - 94 SMT assembly. Your clearly recognize the importance of these topics in your
Average = 55 - 74 assembly processes. This score still offers some opprotunity for improvement.
Below Average < 55 Look at your results on each question and develop an action plan for
improvement if appropriate.

Materials Supply and Validation
1 We evaluate our solder pastes and/or materials with a systematic evaluation procedure, such as "The 12 Board 0
Paste Evaluator" (shown below) or DOE to assure its performance (this question counts 3 times):
2 My engineers have read and understand the solder paste and materials specs and assure they match our use 8
conditions:
3 The response to pause of my solder paste is adequate for my applications: 5
4 The cost of my solder paste and/or materials is not the main criteria for its purchase: 7
5 The printed volume consistency of my solder paste is best of breed: 7
6 My materials supplier(s) understand(s) my process and business needs, we treat each other like partners: 4
7 Few, if any end of line defects can be traced to inadequacies of my solder paste and/or materials: 4
8 The type (i.e. 2, 3, 4, 5) of the solder paste we use matches the application requirement: 4
9 Our organization has a systematic method to assure that the materials/components for future jobs are being 6
prepared while current jobs are being run:
10 Our organization has a systematic method to assure that we have an uninterrupted supply of materials from our 6
vendors:
Ratings:
World Class = > 95 Your score places you below average among users of SMT materials. This
Above Average = 75 - 94
position offers overwhelming opportunities for improvement. Look at your
Average = 55 - 74
results on each question and develop an action plan for improvement. Your
Below Average < 55
organization has an urgent need to recognize that evaluating your solder
pastes and materials is a most important activity.

DOE, SPC, CIP
1 We have a continuous improvement plan that is effective, uses metrics, and is recognized as valuable by the 3
entire organization (this question counts 3 times):
2 We measure our process uptime: 8
3 We measure our unscheduled downtime: 2
4 We measure our line efficiency: 2
5 We measure our work in process time: 5
6 We know our process's Cp and Cpk: 9
7 We have a statistical process control program and use the resulting data effectively to monitor and improve our 9
processes:
8 Our process engineers use designed experiments to optimize our processes and evaluate equipment and 9
materials:
9 Quality is everyone's job: 8
10 Productivity is everyone's job: 8
Ratings:
World Class = > 95
Above Average = 75 - 94 Your score places you as average in DOE, SPC and CIP for SMT assembly.
Average = 55 - 74 This position offers significant opportunities for improvement. Look at your
Below Average < 55 results on each question and develop an action plan for improvement. Having
an effective CIP program is vital for success.

Training and Failure Analysis
1 Our organization has a sophisticated training program for all levels of our staff (this question counts 3 times): 9
2 Our engineers understand the difference between common cause and special cause failures and use this 8
knowledge in process troubleshooting:
3 Our engineers use sophisticated modeling tools, like StencilCoach, Reflow Coach and LineBalancer to help them 5
model processes and perform "what if" analysis:
4 Management uses costing tools like ProfitPro to perform financial "what if" analysis, before making financial 3
investments in equipment etc:
5 Our operators cannot change the process equipment's operating parameters: 8
6 Our engineers know and use analytical problem solving and brainstorming techniques to perform failure analysis: 4
7 There is a process line escalation policy that is understand by all (e.g. if the line is down and remains down this 4
information gets escalated in a documented fashion):
8 Our process engineer's yearly performance review is related to process improvement goals: 4
9 We can perform failure analysis or vend this task out: 6
10 Our staff has all ot the tools necessary to perform their jobs: 6
Ratings:
World Class = > 95
Above Average = 75 - 94 Your score places you as above average in Training and Failure Analysis for
Average = 55 - 74 SMT assembly. This score still offers considerable opportunity for
Below Average < 55 improvement. Look at your results on each question and develop an action
plan for improvement, if appropriate.

Line Level Metrics
Line Efficiency (%) First Pass Yield (%)
Start
60 Goal 98
40 97 99 Goal
80
20 100 96 100
95
94
Start
Line Uptime (%) WIP (hrs)
40 50 60 70 80 90 100 7 6 5 4 3 2 1

But No Matter What
You do, Have a CIP
• Assess yourself
• Establish/Measure Metrics
– Paste Volume
– Productivity Metrics
– Pareto Defects
• Monitor Success/Develop Action
Plan
• Fix the Problems
Software Tools to Help
• In addition to AuditCoach™
• Stencil Coach™
– Helps design stencils including PIP
• WaveCoach™
• LineSimulator™
– Simulates entire line, much easier than
Arena®
• ReflowCoach™

Stencil Design: StencilCoach™
Aspect Ratio
Aperture Width > 4-5
W
particle diameters
Aspect Ratio = W/t > 1.5

t
Area Ratio
Aopening= πD /4
2 Circular Aperture Diameter >
D
8 particle diameters
Awall=πDt =>
t ArR=D/4t ArR = D/4t > 0.66
PW=1-3+ W = PW - AR= W/t

Calculations Recommendations: P/2 2 to 3 > 1.5
Pitch Pad Width PW Aperture W Stencil Aspect AR

Solder
Rectangular Aperatures (P - mils) (PW - mils) OK? (W -mils) OK? Thickness Ratio OK?
Paste Type
(t - mils) (AR)
50 26 TRUE 23 TRUE 6 3.833 TRUE 3
25 15 TRUE 12 TRUE 6 2.000 TRUE 3
20 12 TRUE 10 TRUE 5 2.000 TRUE 3
16 10 TRUE 8 TRUE 5 1.600 TRUE 4
12 8 TRUE 6 TRUE 4 1.500 TRUE 4
35 19 TRUE 17 TRUE 6 2.833 TRUE 3
PW=0-2+ D = PW - ArR=D/4t
Recommendations: P/2 2 to 3 >0.66
Pitch Pad Dia Pad Dia Aperture D Stencil Area ArR
Solder
Circular Aperatures (P - mils) (PD - mils) OK? Diameter OK? Thickness Ratio OK?
Paste Type
(D-mils) (t-mils) (ArR)
40 21 TRUE 19 TRUE 5 0.95 TRUE 3

Pin-in-Paste
Vsolder= 2Vf + Vh - Vp
Vf= Fillet Volume (Pappus-Gu

Vh
Vf Vh= PWB Hole Volume
Vp
Vp= Pin Volume
Inputs Outputs Instructions: T

stencil metrics
Solder Pad Diameter (mils) 76 V = 2Vf +Vh-Vp Cubic Mils component in
Pin diameter mils 18 Vf 17585.147 Cells D32-33 a
PTH diameter mils 30 Vh 40997.784 or stencil ape
PWB Barrel Length mils 58 Vp 14759.202
Paste Reduction Factor 0.52 V 61408.876
Solder Paste Volume Needed 118093.992
If Pin is Square…... Stencil Metrics

Length - mils 18 Stencil Thickness - mils 7
Width - mils 22 Side: If Square Aperture - mils 129.887
Equivalent Pin Dia mils 22.454 Radius: If Round Aperture - mils 73.281
Rectangular? If First Side is (mils): 80
Second Side should be (mils: 210.882

ReflowCoachTM
Time 265
217 C
Temperature Temp 240
Time
Time-s 0 90 140 190 230 250 295 325

Temp-C 30 100 150 209 217 230 217 120
Thruput Calculator
Tunnel Length cm 249 Thruput Bds/min 2.84 Obeys Lee Dwell Criteria? Yes
Belt Speed cm/min 71 Profile Time (min) 3.51
Product Length cm 20
Product Spacing cm 5

Ramp Up to 217 C
250
200
Temperature (C)
150
Lower Limit
Upper Limit
User Profile
100
50
0
0 50 100 150 200 250 300 350 400
Time (seconds)

Time Above Liquidus
260
255
250
Temperature (C)
245
User Profile
Hi Ramp
240 Low Ramp
Temp Max
235 Temp Min
Time Min
Time Max
230
225
220
215
0 20 40 60 80 100
Time (seconds)

Ramp Down
220
210
200
190
Temperature (C)
180
Low Ramp
170 Hi Ramp
User Profile
160
150
140
130
120
290 295 300 305 310 315 320 325 330
Time (Seconds)

The Tools to Do the Job Right
• Statistical Thinking
• DOE
• SPC

Statistical Thinking
Common
Causes Paretos
Special
Histograms
Causes
Control Process
Charts Variation
Ref: Sheri Flori

Statistical Thinking…
The Case of the Squirrel
• The greatest birder
in MA was being
harassed by
squirrels
• It was a crisis
• Even squirrel proof
feeders designed
by NASA wouldn’t
work
• We needed to
apply BMT!
So I Got Serious
A Crossman BB/Pellet Gun

The first shot with BBs.
What should I do?

The second shot.
X
X

The third shot and
fourth.
X X
X
X
What is it telling me?

I try pellets.
X
X X
X
X
What is this telling me?

Accuracy and Precision
• Accuracy: How close
to the measured.
Target Measured
• Precision: How Value Value
repeatable
Accuracy
• Examples of:
– Poor accuracy and
precision (BB’s)
– Good precision, poor
accuracy (pellets w/ bad
sighting) Repeatability
(Precision)

Process Variation = Lost $
LSL USL
LSL USL
LSL USL
Ref: Sheri Flori

Types of Variation
Common Cause Special Cause

• Natural, expected variation • Unnatural, not expected
• Controllable • Possible examples in
• What are examples of CC pellet gun calibration
in calibrating the pellet gun
Knob twiddling can correct neither!

An SMT Electronic Assembly
DOE Example

Random Variation:
Dr. Ron Golf Scores
Number of Rounds Average = 82
Random Variation =
Variance: Sr2
Score

Variation from Factors: Tiger vs Dr. Ron
68 82
Implies that
∆= Difference there is a greater
in averages difference
between Tiger
Number of Rounds
and Dr. Ron,

than among
them ∆2 >> Sr2
STiger SDr. Ron
Tiger’s Scores Dr. Ron

When Variation from Factor Change is Small…….
∆
For example: Phil Mickelson and
David Toms. Then, ∆2 << Sr2
Number of Rounds
Sr
Score

DOE Uses ANOVA
• ANOVA (Analysis of Variance)
– Compares S2 to ∆
• The F Statistic: ∆2
F∝ 2
Sr
• Large F => factors have a significant
effect on result
• “Large” varies with sample size,
typically > 4 for 95% confidence

A Good Stencil Print
*Illustrations courtesy of MPM Corporation

High Speed UFP Printing
• The objective is to perform High
Speed Ultra-Fine Pitch Stencil
Printing.
• Print speed, separation speed and
wipe frequency need to be minimized
to reach this goal.
• Target is 8 second cycle time with
current cycle time >20 seconds

Cycle Time Breakdown
10
21.8 sec.
0
LOAD VISION PRINT SEPARATE UNLOAD Wipe
• 10 inch wide board printed at 1 inch

• 2 s Load + 2 s Vision + 10 s Print + 5 s Separate + 1.8 s Unload + 1 s Wipe = 21.8 sec

Cycle Time Breakdown
10
0
LOAD VISION PRINT SEP. UNLOAD Wipe
Objective:
1.8 s Load + 1.8 s Vision + Print + Separate + 1.5 s Unload + Wipe < 8sec

Factors
• A: Print Speed: 1=4, 2=6, 3=8 inch/sec
• B: Separation Speed: 1 = 1, 2 = 1.5 sec Number of Runs:
• C: Wipe Freq: 1= 1/8, 2= 1/12 3x2x2x2x2=48
– 1/8 adds 0.875 sec, 1/12 adds 0.6 secs Not that many!
• D: Stencil Type: 1= Efab, 2 = Laser
• E: Paste: Vendor A =1, Vendor B =2
• Full Factorial
• Assume no interactions
• Aperture: W= 8, t= 5, L= 64 mils=>2560 mils3
• Desire paste volume to be +/- 10% of aperture
volume
• Response: Solder Paste Volume

MiniTab Results
Source DF Seq SS Adj SS Adj MS F P
Print Sp 2 175117 175117 87558 2.48 0.096
Separati 1 104533 104533 104533 2.97 0.093
Paste Ty 1 559008 559008 559008 15.86 0.000
Stencil 1 274519 274519 274519 7.79 0.008
Wipe Fre 1 902008 902008 902008 25.60 0.000
Error 41 1444731 1444731 35237
Total 47 3459917
For stencil and paste, we can reject H0 Minitab is available
with confidence. Hence, we can select Free for 30 days at
the better choice for each of these factors
and re-experiment to optimize total speed.
Minitab.com!

SPC and Variables Data
• Data relating to a specific process
step
• Quantitative
• Can be used to monitor and improve
process performance
• Example: Solder Paste Volume
• Variables data are crucial for an
effective SPC Plan
Process Data: Attribute Data
• Data that relates to the 350
300
performance of the 250
product 200
• Examples: 150
100
– Shorts 50
– Opens 0
Shorts Opens Missing Solder Balls
Compnent
– Missing Component
• Important, but cannot be Pareto Attribute Data
measured to improve
process performance
• Any plan should strive to
relate attribute data to
variables data and develop
a CIP around this
relationship

Variables Data Example
Opens
Shorts
Upper Spec Limit Lower Spec Limit
Solder Paste Volume Solder Paste Volume
= Average

Variables Data: Solder Paste
Volume
Process Occurrences
LCL UCL
LSL USL
Solder Paste Volume

Control Chart for SPC
USL
UCL
Sigma 3
Sigma 2
Sigma 1 Centerline
Sigma 1
Sigma 2
Sigma 3
LCL
LSL

Solder Paste A
Solder Paste B

Paste B Process Capability Sixpack for Print Volume
Xbar and R Chart Capability Histogram

1
1900 5 UCL=1898
1800
Mean
Mean=1750
1700
5
1600 LCL=1602
1200 1700 2200
Subgr 0 100 200
600 Normal Prob Plot
UCL=541.0
400
Range
R=255.9
200
22
0 LCL=0
1500 1750 2000
Last 25 Subgroups Within Capability Plot

1950 StDev: 110 Process Tolerance
Cp: 1.06 Within
1800 Cpk: 1.06 I I I
Values
Overall
Overall I I I
1650
StDev: 111.634 Specifications
Pp: 1.05 I I
1500 Ppk: 1.05 1400 2100
180 190 200
Subgroup Number

Paste A Process Capability Sixpack for Print Volume
Xbar and R Chart Capability Histogram

2050 1
1 1 1
8 1 8 5 UCL=1898
1800
Mean=1750
Mean
2
66 6 6 22 2 66 6 62 6 66 2222
1550 6 51 5 56 55 5222222225 6151 5 5 5622 1 6 5 2666 2 52 55 562 LCL=1602
11 111 1 111 11 111 1 1 1 11111 1 1111 1 11 1 1 1 11 11
1 11 1 11 11 1 1
1 1
1 1 1 1 1 11
1 1 1
1300 1
500 1500 2500
Subgr 0 100 200
1
Normal Prob Plot
1500 1 1
11 1
1 1 1 1 1 11 1 1
11 1 1 1
1000 1
1
1 111 11 1 1 1 1 1 11
Range
1
1 11 1 11 11 111 11 11 1
1
1 1 1 11 1 111 1 1 11 11 1 111 11
1 1 1 11 1 1 1 111 111 11 1 1 1 1
1 111 2 1 1 11 11111 121 2 111 1 11 12 1 1 11 1 21 2 2 121 1 2 2 2 11 2
2
500 22 2 2 2 22 222 2 2 2 2 2 22222 2 22 2 22 2 UCL=541.0
2 2 2 22 2 22 2
2 2 2 2 2 222
2 2 22 2 22 2 2
R=255.9
0 LCL=0
600 1600 2600
Last 25 Subgroups Within Capability Plot

2500 StDev: 110 Process Tolerance
Cp: 1.06 Within
2000 Cpk: 1.06 I I I
Values
Overall
Overall I I I
1500
StDev: 273.152 Specifications
Pp: 0.43 I I
1000 Ppk: 0.43 1400 2100
180 190 200
Subgroup Number

Best Practices Summary
• Assess Yourself
• Use Metrics
• Develop and Implement Action Plan
• Monitor Success
• Continuously Improve

Pb-Free Board/Component
Finishes
Board Finishes Component Finishes
– Immersion Ag – Sn/Pb
– OSP – 100% Sn
– Electroless Ni/Immersion – Pd/Ag
Au – Ni/Pd
– Immersion Sn – Ni/Sn
– Ni/Au
– Ni/Pd/AU
Pb Contamination and ability to withstand the higher

reflow temperatures are the main concerns.

Component Finishes: Pure Sn
• Easiest and most obvious choice
• Sn Whiskers still a major concern
• So TI recommends Ni/Pd/Au

Structure of Pb-Free Preplated Finishes
NiPdAu finish: Present

NiPd finish: Past
GOLD: 30-150 ANGSTROMS
PALLADIUM: MINIMUM OF 3u” PALLADIUM: 0.2-6.0 u”
NICKEL: 40 - 60 u” NICKEL: 20 - 80 u”
Pd / Ni STRIKE: < 5 u”
Ni STRIKE: < 5 u”
COPPER BASE METAL COPPER BASE METAL
• NiPdAu structure shown has been in use since early 1990s.

• Enhanced wetting performance with NiPdAu finish seen in solderability tests.
• See TI Application Note SZZA026 for complete evaluation of NiPdAu finish.
Courtesy: James Huckabee, TI

NiPdAu (SOP) Board Mount: Visual Appearance
Typical wetting NiPdAu finished SOP Typical wetting NiPdAu finished SOP
components with SnPbAg solder, NiAu components with SnAgCu solder, NiAu
PWB finish. PWB finish.
Visual Appearance Results: All solder joints exhibited a heel fillet height at least one times the lead thickness and
evidence of wetting to the sides of the leads. This performance would be considered acceptable for all 3 classes of
products identified in IPC-A-610C (general electronic products, dedicated service electronic products, and high
performance electronic products.
Courtesy: James Huckabee, TI

PWB Surface Finish Summary
HASL ENIG OSP ImSn ImAg
Thickness Au: 3 - 8
100 – 1000 8 - 20 40 – 60 3 - 12
(microinches) Ni 50 - 150
Fine pitch
Poor Excellent Excellent Excellent Excellent
quality
Contact Not
Fair Good Good Good
Connections Recommended
Not Not Not

Wire Bonding Limited Limited
Recommended Recommended Recommended
Cost
1X 2X 0.3X 1X 1X
(To HASL)
Availability High Moderate High Very Limited Limited
Hazard to
High Moderate Low High Low
Manufacture

2006 Finish Estimations
Additional slides available at www.pb-free.com Projections courtesy of Enthone, Cookson Electronics

Immersion Ag
• Low Cost
• Planar Surface
• Compatible with touchpads /
solderless connections (if thicker Ag
is used)
• 2 – 10 microinches typical thickness
Additional slides available at www.pb-free.com

ImAg Issue #1: Tarnish
• Just like Ag silverware, this finish
will yellow over time
– Will occur during assembly
– Often, purely cosmetic
• Solderless connections appear very
tolerant of tarnish
• Thicker Ag less prone to tarnish

ImAg Issue #2: Migration
• The presence of moisture and
current can cause the migration of
Ag ions
• Migrate from cathode to anode
forming dendrites that reduce
resistance…may eventually form a
short

ImAg Issue #2: Migration
NO SIGN OF DENDRITIC GROWTH

*graph from www.alphametals.com

ImAg Issue #3: Premature
Intermetallic Failure

ImAg Issue #3: Premature
Intermetallic Failure
• Thick Ag finishes tarnish less and
are more compatible with solderless
connections, but…
– Thick Ag means more organic co-deposit
– Organic co-deposit must be forced out of
molten solder
– Non-expulsion of organics can result in
microvoids along board/solder intermetallic

Pb-Free Solder Paste Evaluation:
Pre-Screening
Since there are 10-20 solder paste
suppliers, it is essential to narrow down the
Pb-free candidates prior to actual physical
evaluation

Lot-to-Lot Viscosity Stability

Typical Pb Contamination
Pb Content Must be less than 0.05%

Pb-Free Reliability
• Surface Insulation Resistance
– Test procedure in J-STD-004
– 7-day test to determine if flux residue with
affect electrical reliability
• Electromigration
– Test procedure in Telcordia GR-78
– 21-day test for electrochemical migration

Pb-Free Supplier Support
• Assess your company needs against
supplier capabilities.
– Are they available globally?
– Are they available locally?
– Can they accommodate changing
requirements?

Pb-Free Solder Paste Evaluation:
Printing through Reflow

Approximate Distribution of
Process Related Defects
Component Placement
15%
Reflow
15%
Incoming Components Solder Paste Screen Printer

6% 64%

Evaluating the Pb-Free Print Process:
Four Main Paste Variables
• Viscosity Relative to Production
Temperature
• Stencil Life
• Response to Pause
• Resistance to Excessive Shear
Thinning

Effects of Viscosity
Too High Too Low
• Poor rolling • Excessive deposit

• Blade Hang-up • Solder balling
• Aperture clogging • Solder beading
• Insufficients • Slumping
• Bridging

Viscosity Vs. Temperature

Solder Paste Stencil Life
Tack Life
3
Delta Tack Force (grams/mm2)
0
0 5 10 15 20 25 30 Brand X
-1
Good Paste
-2
-3
-4
-5
-6
Time (hrs)
Inspired by: MPM

Effects of Short Stencil Life
• Variability in solder paste deposits
• Loss of “up time”
• Wasted paste, wasted money

Solder Paste Response to Pause
One Hour Pause
Response to Pause
50
45
Volume of Print (cubic mils)
40
35
30
Brand X
25
Good Paste
20
15
10
0
0 2 4 6 8 10 12 14 16 18
Print Number Inspired by: MPM

Effects of Poor Response to
Pause
• Loss of up time
• Paste deposition variability
• Higher cost of “time for assists”

The Twelve Board Paste Evaluator
1. Start with enough paste for 12 prints
2. Print 4 boards (no kneading)
Metrics to Measure:
• Print Volume
• Print Definition
• Volume and Definition after Idle
• Release from Aperture
• Squeegee Hang up
Two hour sit, place, Six hour sit, place, • Tack
measure tack measure tack • Solder Joint Quality
For Final Candidates
•Coalescence
1 Board, 1 hr sit 1 Board, 3 hr sit •Reflow Window
then reflow Repeat •J Standards
then reflow
3. Pause one hour, no kneading, print 4 more

boards, repeat tests in 2
4. Pause one hour, no kneading, print 4 more
boards, repeat tests in 2
The above is good, but should also test for shear thinning!

Solder Paste Resistance To
Shear Thinning
50
40
Acceptable
Tack (grams)
30
20
Unacceptable
10
0
0 60 120 180 240
Time (minutes)

Effects of Excessive Shear
Thinning
• Excessive deposit volume
• Slumping
• Bridging
• Balling/Beading

Measuring Shear Thinning Effect
• Often overlooked in solder paste evaluations.
• Discovered once a solder paste is implemented into
production (too late).
• Requires many print strokes, thus many boards to
discover.
• One suggested technique:
• 1. Place fresh paste onto stencil (Repeat for all pastes being evaluated)
• 2. Set printer to run 30 knead strokes, wipe the underside of the stencil.
• 3. Print one board
• 8. Measure response variables on each printed board

Placement Capabilities:
The Solder Paste Tack
• Should not be much different than
standard Sn/Pb solder pastes
• Suggested evaluation technique:
1. Using fresh paste, print three boards
2. After 8 hours, place components onto one board
3. After 24 hours, place components onto the second board
4. After 48 hours, place components onto the third board
5. Compare the results and determine which solder paste
lost more components, during the placement process, at
each time interval

Evaluating Pb-Free Reflow
• Most analyzed aspect of transition to pb-
free.
• Primary Pb-free reflow response variables:
– Wetting
– Appearance
– Voiding
– Solder Balling
– Tombstoning
• Critical Evaluation Criteria:
– Time above liquidus (TAL)
– Peak temperature
– Soak time

Pb-Free Profile Variables
Reflow Profile for Indalloy #241
95.5Sn/3.8Ag/0.7Cu
300
250
PEAK
MP = ~217 C
Temperature (C)
200
150
100
Soak Zone TAL
50
0
0 1 2 3 4 5
Time (Min)

Reflow DOE
Low Middle High
Soak Time 1 minute 2 minutes 3 minutes
TAL 40 Sec. 60 Sec. 80 Sec.
Peak Temp. 230 °C 240 °C 250 °C
Are 27 profiles realistic with multiple solder pastes?

Categorizing the Response
Variables
• Potential defects need to be ranked
according to each individual process.
– Which is more critical to the process?
• Probe testability
• Residue color and quantity
• Etc.
– Which is more detrimental to the product?
• Voiding
• Tombstoning
• Bridging
• Etc.

Analyzing the Overall Results
• 60 – 70% of all defects are attributed
to the stencil printing process.
– Should be considered the most important
for overall process consideration
• Reflow is a new crucial variable for
Pb-Free
– Not necessarily an issue for paste
– Critical for components and boards

Paste Evaluation Summary
• Pb-free transition requires a knowledge of
statistics and DOE to have a successful
implementation.
• Pre-screening of solder pastes necessary
to make evaluation practical in size.
• Printing and reflow require careful
analysis to adequately distinguish
between solder pastes.

Motorola Pb-Free Cellphone
Assembly
Courtesy: Vahid Goudarzi,
Motorola

Motorola’s Process Criteria
• Paste must have good response to pause,
tack, slump and other printing metrics
• The process/paste must show good
coalescence and solder joint quality in a
broad reflow process window
• The reliability of the finished product must
be as good or better than the standard Pb
solder
• The process must be simple and robust so
that it can be transferred to other
locations world wide

Paste Evaluation/Manufacturing
Process Development
• Screen Printing Evaluation
• Reflow Profile Development
• Tackiness Measurement
• Surface Insulation Resistance
(SIR) Evaluation

Pb-free Solder Paste Evaluation
Paste evaluation & selection strategy:
Work with 8 preferred paste suppliers to develop a lead-free solder paste that
meets Motorola’s manufacturing quality & product reliability requirements
The Finalists: Flux Vehicles

Phase # 1 Phase # 2 Phase # 3
A A1 A2 A3
Paste B1 B2 B3
B
Suppliers
C C1 C2
Lead-free solder paste suppliers & materials
These studies were completed using Sn/Ag/Cu, Entek finish

boards, & air atmosphere

Stencil Printing Evaluation
• Objective: To ensure Pb-free paste
performs consistently as a function of
time
• Variables:
– Abandon time @ t=0, t=1, & t=4 hours
– Solder paste (A1, B1, C1, A2, B2, C2, A3, B3)
• Output:
– Volume measurement using laser system
– Visually inspect for smearing and selected
apertures for clogs.
Optimum Print Speed, Squeegee Pressure, & Snap Off was set per paste
supplier recommendation and validated

OSP Finish Test Vehicle for Paste Evaluation
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 171 8 1 9 2 0 2122 23 24
R
S
C
A
B
C
D
E
F
C
S
R
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 171 81 9 20 21 22 23 24
Selected Inspection Sites

Based on gage R&R
results

Solder Paste Volumetric Measurement
or 12 mils SMD pads @ t=0 h
400
Volume
300
200
Paste C1 Paste C2 Paste B3 Paste B1 Control Paste A1 All Pairs
Tukey- Kramer
0.05
Pb-free Solder pastes performed well @ abandon time = 0h

Solder Paste Volumetric Measurement
for 12 mils SMD pads @ t=1 h
400
300
Volume
200
100
Paste B1 Paste A1 All Pairs

Paste C1 Paste C2 Paste B3 Control Tukey-Kramer
0.05
Paste C2 failed @ abandon time=1h

Reflow Profile Development
• Objective: To determine reflow process
window & identify a Pb- free paste which
requires MINIMUM peak temp.
• Variables:
– Peak temperature
– Time above liquidus
– Solder Paste (A1, B1, C1, A2, B2, C2, A3, B3)
• Output:
– Coalescent performance
– Solder joint quality

Reflow Profile Matrix
Pb-free 2X3 Full Factorial Reflow DOE
• Minimize peak temp. to reduce thermal stress on Components
• Interaction between peak temp. & time above liquidus
Time Above Liquidus
60Sec. 70Sec. 80Sec.
Peak Temperature
Peak temp
229C P1 P4 P7 Ramp rate
237C P2 P5 P8
Time above
liquidus
245C P3 P6 P9
Lead-free reflow profile
Selected paste MUST perform equally well @ P1 through P9 in
air atmosphere

Reflow Profile Development Cont.
Inspection criterion:
Coalescent performance @ P1,P2, P3, P4, P5, P6, P7, P8 &P9
Poor Coalescent Good Coalescent

Poor coalescent is attributed to powder oxidation during reflow
process in air atmosphere

Reflow Profile Development Cont.
Inspection criterion:
Wetting performance @ P1,P2, P3, P4, P5, P6, P7, P8 &P9
Insufficient toe fillet
Poor Solder Joint Good Solder Joint
Insufficient toe fillet results in field reliability issues

Flux Tackiness Measurement
• Objective: To ensure flux provides sufficient
tackiness to hold components in place during
manufacturing processes
• Variables:
– Paste life @ t=0;t=1h t=2h; t=4h; t=8h
– Pb-free solder pastes
• Output:
– IPC-TM- 650 Test Procedure: Measure the force
required to Separate a 5mm diameter probe from paste
– Shake Test -Automated vision inspection after
placement

Flux Tackiness Measurement Cont.
IPC650 Tack Test - Tack test evaluation result
Tack Test
Tack Test
2.5
2.5
2
Tack (g/mm2)
2
Tack(g/mm2)
Control
Indium 51A #1
1.51.5 Indium 92J
Control #2
Indium 232-99-2
1 B1
Indium SMQ 230
1
0.5 B3
0.5 0
0 1 2 4 8
Time (Hrs)
0
0 1 2 4 8
Time (Hrs)

Flux Tackiness Measurement Cont.
To ensure flux provides sufficient tackiness to hold
components in place during assembly process
Component placement offset after 120 second of
shaking by chip shooter
1) Populated PCBs after 0, 4, and 8
hours
resistor
-3
1
2) Image components to determine
X, Y, and Theta offsets. 3.2x1.6
-1
-1
0
3) Place PCBs on XY table of Chip caps
1
3
0
Shooter & shake PCBs for 120
tantalum
1
2
1
Sec.
switch
-1
0
4) Image components to determine
X, Y, and Theta offsets caps
1
1
5) Determine delta for before & after -4 -2 0 2 4 6
shake process X-Offset Y-Offset Theta

Surface Insulation Resistance
Motorola SIR Test Boards = B25 Test Board + Solder mask
Supplier B Surface Insulation Resistance Test
109
Channel 1
Ohms
Channel 11
Channel 21
Channel 153
107 Channel 163
24 72 120 168 216 264 312 360 408 456

Hours
SIR requirements is minimum of 10 8 Ohms

Paste Final Evaluation Results Phase 3
Phase 1 Phase 2
Paste Paste Paste Paste Paste Paste Paste Paste
B1 A1 C1 B2 A2 C2 B3 A3
Visual Inspection
Printing Paste
Volumetric data NT NT NT
P1,P5,P9
Reflow
P1,P2,P3,P4,P5,
P6,P7,P8,&P9
Instron IPC650 NT NT NT NT NT NT
Tackiness
Shake Test NT NT NT NT
Solder Joint NT NT NT NT NT
Quality
ALT NT NT NT NT NT NT
J-STD B25 NT NT
SIR
Motorola NT NT NT NT NT
Passed Failed NT Not tested

Paste B3 met all requirements

Paste Final Evaluation Results Cont.
Solder Coalescent Comparison @ P1, P5, & P9
Supplier A Paste Supplier B paste
B3 B3
A1 A1 B3 B3B3
@ @
@ @ @ @@
P1 P5 P1
P1 P5
P5
A1 B3
B3
@ @
P9 P9
Poor Coalescent Good Coalescent

Paste A1 does not fully coalesce and result in grainy joint due to powder
oxidation in air atmosphere

Solder Joint Evaluation @ P1, P5, & P9
Lead-free @ P1(229;60) Lead-free @ P5(237;70)
Lead-free@ P9(245;80) Leaded @ 210 C

No significant difference in solder joint fillet @ P1, P5, & P9 using
B3 solder paste

Solder Joint Evaluation @ P1, P5, & P9
B3 B3
@ @
P1 P5
Insufficient toe
fillet
B3
Sn/Pb
@
P9 Paste A1
No significant difference in solder joint fillet @ P1, P5, & P9 using

B3 solder paste

Intermetallic formation
P9
0.0025 mm
P1 P5
0.0023 mm 0.0025 mm

Recommened 235C profile
Recommended Profile for B3
300
Time above 217C: 70s Peak 235C
250
200
Temp/C
150
Ramp
100
0.7
50 deg/sec
0
0 1 2 3 4 5 6 7 8
Time/min
Peak Temp. = 235 C +/- 5C; Time Above Liquidus = 70Sec +/ 10Sec

Reliability Evaluation
Product Level & Solder Joint Reliability Evaluation
• Drop Test
• Shear Test
• Liquid-to-Liquid Thermal Shock
• ALT for different Products
Pb free solder joints MUST perform equal or better

than leaded solder joints

Reliability Evaluation Cont.
Drop Test 25
% of crack
1) Dropping products from 5 feet 20
Leaded radios
15
2) Vert. & horiz.vibration for 2 hrs 10
3) Thermal shock for 48 hrs 5
4) Repeated step 1 thru. 3 X times 0
5) Measure % joint cracks on shields
7
e2
6
4
1
5
3
ld
ld
ld
ld
ld
ld
ld
ie
ie
ie
ie
ie
ie
ie
Sh
Sh
Sh
Sh
Sh
Sh
Sh
25
20
% of crack
Pb-free radios
Drop test vehicle
Shields
15
10
5
0
4
1
7
Sh 6
3
2
5
ld
ld
ld
ld
ld
ld
ld
ie
ie
ie
ie
ie
ie
ie
Sh
Sh
Sh
Sh
Sh
Sh
Shield solder joint cracking is significantly reduced using B3

Solder Joint Reliability Evaluation
Test Vehicle SOIC
0.5mm QFP
0.5mm CSP
• 6X6 mm Package size CSP20X40 Cap.
• 0.5 mm pitch partial array
• 0.3 mm solder balls size 0.8mm
CSP
0.5mm
Conn.
BGA
0.75mm CSP
0.5 mm CSP
DIME

Shear Test
•10
•SnPb
•9
•SnAgCu
•8
•Shear at failure (kg)
•[after thermal shock]
•7
•6
•5
•4
•3
•2
•1
•0
•Ceramic •Tantalum •Small •Ferrite Bead •Mid-size

•Inductors •Capacitors •Capacitors •capacitors
No significant difference in shear force after LLTS.

Liquid-to-Liquid thermal shock evaluation
(-55 °C to +125 °C)
Variables:
- Solder Paste (Paste B3 & Pb Paste)
- Component Type ( 0402, 0603, 0805, BGAs, CSPs, VCO,
Transformer)
Output:
- Electrical test at every 75 cycles for 450 cycles
- Red dye analysis at 150, 300, and 450 cycles

Liquid-to-Liquid thermal shock results after
450 cycles
Joint crack data for different components
125 Passed Joint
100
Crack area, %
75
50
Failed Joint
25
0
Failed Joint
-25
Pb-free Sn-Pb All Pairs
Tukey-Kramer
Solder 0.05 Red dye evaluation result
No significant difference in cracked area in leaded and Pb-free joints

Products Built with Pb-free Paste
* Concorde iDEN * i700 iDEN
* i1000 iDEN * i85 iDEN

* i1000 Charger
Products built with Pb-free solder paste and passed ALT

Manufacturing Issues
• Tombstone Failures
• Air Voids on CSPs
• Logistics: Have a Plan to Avoid
mixing SnPb and Pb-free Assembly

Manufacturing Issues Cont.
20X40 Tombstone Failures
Lead-free Solder paste is more prone to tombstone failures due to
higher coalescent force.
T3 T3
T1 T2
T4
T5
Before Reflow After Reflow

T1 & T2 : Tack Force
T3 : Weight
T4 : Surface Tension (outside)
T5 : Surface Tension (underneath)
T4 is significantly higher using lead-free solder paste

20X40 Tombstone Failures
Tombstone failures are attributed to lead-free solder paste & blind vias
Blind Via
Pads without blind vias did NOT show tombstone failures

20X40 Pad Design for Conventional PCB
Evaluated Circle, ½ Circle, Rectangular (vertical & horizontal),

Oblong, square stencil, etc.
0402 Stencil Aperture Openings
.015" .011" .018"

Stencil design to minimize .007"
tombstone failures C .022"
A*
.041" .008"
on pads with blind vias
.047"

Tombstone / floating Failures
Paste volume was reduced to eliminate tombstone failures on large

discrete inductors

Manufacturing Issues Cont.Voids
Air voids on CSPs
• 6X6 mm Package size
• 0.5 mm pitch partial array
• 0.3 mm solder balls size
0.5 mm CSP
DIME

Voids Mechanism in CSPs
1) Solder bump oxidation
2) Flux out gassing

Air voids on CSPs
BGA/CSPs are more prone to voids mainly due to leaded bumps on package &
increased oxidation of powder due to higher reflow temp.
Solder Pastes
Variables:
Paste B3 Paste #15 Paste #16
- Ramp Rate
- Solder Paste Ramp Rate (Deg./Sec.) 0.5
CSPs CSPs CSPs
Outputs:
- Number of voids 0.8 CSPs CSPs CSPs
- Void size
1.5 CSPs CSPs CSPs

Air Voids in CSPs joints
Void Quantity VS. Ramp Rate Void Size VS. Ramp Rate
25
30
20
Average Void Size (% of joint)

25
15
20
10
15
5
10
All Pairs All Pairs

0.5 Deg/Sec 0.8 Deg/Sec 1.5 Deg/Sec 0.5 Deg/sec 0.8 Deg/sec 1.5 Deg/sec
Tukey-Kramer Tukey-Kramer
Ramp Rate 0.05 _ID_ 0.05
• Quantity of voids are not significantly affected by ramp rate

• B3 shows significant reduction in void size as ramp rate increases

Pb-free Joint&Component Appearance
4) Joint & Component Appearance
Coalescent Performance Comparison
Pb
PbPaste
PasteAfter
BeforeReflow
Reflow Pb
Pbfree
freePaste
PasteBefore
After Reflow
Pb paste fused onto Cu coupon Pbfree

freepaste
pasteprinted
fused onto
Pbpaste
Pb Before Reflow
printed onto Cu coupon Pb ontoCu
Cucoupon
coupon
Pb-free Paste has a significantly higher Coalescent force

Solder Joint Comparison
PasteB3 Control(Pb)

Pb-free Joint &Component Appearance Cont.
Shield Discoloration
SnO & SnO2 is formed after lead-free reflow process.
Leaded reflow profile Lead-free reflow profile

Oxidation does not affect electrical performance

Motorola Pb-Free Implementation Summary
9 Solder Paste Selection - Evaluated 19 different Pb-free solder pastes
and selected B3 based on manufacturing and product level reliability
requirements.
9 Manufacturing processes - Reflow profile, screen printing operation,

tackiness evaluation, etc. completed
9Reliability Evaluation - Pb-free solder joint reliability evaluation has

shown equal or better performance compared to current materials
9Components 100% of the components Pb-free qualified

-
9Electrical & Mechanical 100% completed with NO issues

-
9Quality - No manufacturing/product quality issues; DPHU goal were met

Motorola Pb-Free Summary
•Production since 09/04/01
•Many site implementation
•More than 1M cell phones have been

shipped to the field
• No field reliability issues have been

encountered

Wave Soldering and Large Board
Issues

Lead-free Complex Board
• Functional units were assembled using current materials,
equipment set and a no clean Pb-free solder paste
mid-range server board Pb-free thermal profile

• materials • linear heating ramp – 0.9 °C/s
– surface finish: Ni/Au • average peak temperature - 247
– board resin: hi Tg FR-4 resin °C
– current component • dwell time – 75s above 217 °C
technologies • Courtesy: Eddie Hernandez, HP

Large Board Summary
• Higher delta T on board => higher
peak T = 245oC
– Reflow profiling and control much more
important
• DOE needed for Process
Optimization
• NiAu and OSP were successful
• Similar manufacturing issues to
Motorola
Wave Soldering
• Must use SAC, not SnCu
– SnCu does not process well and can cause
Tin Pest
• Process Control is crucial

Transformation of Beta-Tin into
Alpha-Tin in Sn-0.5Cu at T <10oC
Ref: Y. Karlya, C. Gagg, and W.J. Plumbridge, “Tin pest in lead free solders”, Soldering and Surface
Mount Technology, 13/1 [2000] 39-40

Wave Soldering Overview
Exhaust
Preheating (IR)
Conveyor
Cooling
Fan
Preheating (CVX)
Fluxer Air Knife
Inspired by Steve Breed, Speedline Chip Wave Laminar Wave

The Process
• Angle of Conveyor: • Wave Form
Typically fixed at 6o – Don’t use chip wave if you
• Conveyor Speed: don’t have chips!
•
– Thermal Mass Dependent Depth of Immersion: 50%
•
– 150 cm/min typical Solder Purity: Monitor
• Flux: According to Specs! Monthly
• Preheat parameters: • Solder Temperature:
– Too hot =>drives off flux 260oC
– Too cool => no activation
– Temp Rqmts depends on flux
type

Adequate Fluxing is Crucial
• Follow the flux and
sprayer
specifications to
assure coverage
with in spec
• Measure coverage
with a fluxometer
http://www.ecd.com/emfg/instruments/fluxometer/index.asp

“KIC”: PWB Temp Profile
Board Bottom
183 C
∆T<140oC
Temperature
Board Top
Ramp Rate: 2-4oC/s
Entrance of Wave: Time

Wave
>100oC Alcohol Flux
>120oC Water Flux

WaveCoach™ can Help
Board Bottom
183 C
Temperature
Board Top
Time
Time (sec) 0 90 150 170 200 Conveyour Speed (m/min)

Temp Top -C 30 70 90 Length of Waves (cm)
Temp Bottom 120 235 160
Results
Time in Wave (s) 3.86 PWB Bottom ∆T 115.00 Cool Down Rate (C/s) 2.50 PWB Bottom T at Wave 120.00 PWB Top T at Wave
Within Spec? FALSE FALSE TRUE TRUE

The Contour Wave Form
PWB pulls solder over exit
Courtesy: Electrovert

Calculating:
The Dwell Time
• Dwell time (DT) related to conveyor
speed (CS) and “length” of wave
(LW).
– DT = LW/CS
• LW = 3 “, CS= 5’/min
– What is DT?
• Use LevCheck™
– Glass with grid
Optimizing the Process
1. Setup the Fluxer 4. Set depth of
2. Establish pre-heat immersion at 50%
profile 5. Monitor solder
3. Set the contour purity monthly
wave to just pull • Alpha “Pot Rite”
solder over the program
exit wing 6. Set Solder
• Only use chip wave Temperature
if you have chips 7. Set Dwell Time

Thanks for Coming!

Best Practices Factory

Uploaded by

Copyright:

Available Formats

Best Practices Factory

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Best Practices Factory

Uploaded by

Copyright:

Available Formats

Implementing Pb-Free

Assembly at Your Factory

Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

All slides available at www.pb-free.com

Slide #1 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Slide #2 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Slide #3 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

All assembled on 30-35K lines!

Slide #4 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Slide #5 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Assembly & Test: 14%

Slide #6 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Source: The Internet

Slide #7 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Die Attach Adhesive Courtesy:

The silicon die is the heart of electronics, it produces all of

Slide #8 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Slide #9 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Stencil Chip Shooter Pick and

Hand Mount and Opto

Slide #10 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

• Japanese Home Electronics

• USA? Some state activity, nothing

Slide #12 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Slide #13 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Slide #14 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Slide #15 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

MJ70 Minidisc Player MJ30 Minidisc Player

Slide #16 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

H845L Camcorder PC Audio Board, Flora 220CX

Slide #17 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

– Batteries: 4,000,000 tons

Slide #19 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

• Paint and gasoline reductions

Slide #20 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Slide #21 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Slide #22 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Slide #23 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Slide #24 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Despite the data, Pb-free solder in the Electronics

Slide #25 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Slide #27 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Slide #28 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Slide #29 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Slide #30 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Slide #31 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Slide #32 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Slide #33 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Slide #34 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Slide #35 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Slide #36 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Slide #37 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

• Phase 1: DOE to analyze 10 probable

Slide #38 Pb-Free Workshop © Ronald C. Lasky, Timothy Jensen 2003

Alloy Solidus Liquidus

63Sn37Pb 182.1 183.0