Providing a Strategic Technical Advantage
and Corporate Partnership
About Universal Instruments
• Universal Instruments drives the advancement of
the global electronics assembly industry by
supplying first-class equipment and process
expertise. Universal’s binding quality policy and
culture of high integrity and ambition provides a
strong foundation for long-lasting productivity and
profitability.
▫
▫
▫
▫
▫
▫
▫
▫
Founded in 1919 in Endicott, NY
Serving the electronics industry since the 1950s
Global company with regional infrastructure
Broad customer base across all regions and tiers
Solution provider for all placement applications
Strong brands and broad market position
Installed base of 20,000+ systems
Proven technology leader: original SM Platform
concept, over 180 industry patents
page 2
Universal Instruments’
Infrastructure
• 5 Product Lines
▫
▫
▫
▫
▫
page 3
Surface Mount
Insertion Mount
Advanced Packaging
Automation
Advanced Process Laboratory
Advanced Process Laboratory
Binghamton, NY
• Founded in 1987
▫ First and most complete Advanced Process
Laboratory in the industry generally accessible
to the public.
• APL is made up of 3 interactive groups
▫ Consortium - Capstone to Our Knowledge
▫ Process Support
▫ Analytical Services
• Consortium founded in 1992
▫ Research into surface mount assembly and TAB
• Complete Analytical Laboratory
• Full process development and production
capability
▫ ITAR Compliant & ISO 9001 Certified
page 4
Consortium
Process
Support
Analytical
Services
The APL Capability
http://www.uic-apl.com/uic-apl-equipment-list
▫ Assembly
▫ Analytical Characterization
Production
Pr
roduction Simulation
Simulation
▫ Environmental Testing
▫ 8850 ft2
Analytical
A
nalytical
L
aborattor y
Laboratory
Metallurgical
M
etallurgical Prep
Prep
Environmental
Envi
ironmental Testing
Testing
Tes
page 5
Mechanical
Laboratory
AREA - Research
http://www.uic-apl.com/Research-Plans
Project
Classification
Materials
1
2
3
4
5
6
7
8
9
10
11
Reliability
12
13
14
15
16
17
18
19
Advanced
Process
page 6
20
21
22
23
24
25
26
27
MAT1A. Underfills and Adhesives
MAT2A. Circuit Board Materials
MAT2B. Pad Cratering Dependence on Glass Type
MAT3A. PCB Surface Finishes
MAT4A. Thermal Interface Materials
MAT5A. Paste and Flux Evaluations
MAT6B. LF Die attach and New Alloys
MAT6C. Shear and Fatigue Testing of High Temperature Solder Alloys
MAT7A. New Lead-Free Solder Alloy Evaluations & Microstructure
MAT7B. Effect of Reflow Profile on Mechanical Properties of Various Solder Alloys
MAT7C. Effect of Precipitate Size and Spacing on Thermal Fatigue Performance on LF
Solder Joints
MAT8A. Impact of Conformal Coating on Thermal Cycling Reliability of SMT Components
REL1A. Prestress & Pad Cratering
REL2A. Lead-Free Phenomenological Model
REL3A. Vibration Testing Methodologies
REL4A. Creep Corrosion
REL6A. Print Correlations to Reliability
REL9A. PCB HDI Robustness
REL10A. Drop test JESD22-B111 Redesign Evaluation Characterization of Proposed
JEDEC Drop Test Vehicle
REL11A. Compression of Second Level Interconnections and the Effect on ATC Reliability
APD1A. Broad Band Printing Process
APD3A. Advanced Packaging Considerations
APD3B. WLCSP RDL Reliability
APD4B. Rework of MLF Devices
APD6A. AOI/SPI Defect Detection (0201)
APD7A. Hand Soldering Process for High Temperature Electronics
APD8A. LGA/BGA drop test reliability
Co-PI
Prof. Co-PI
PI (UIC)
Kondos Mohammad Borgesen
Kondos Mohammad Borgesen
Kondos Mohammad Borgesen
Kondos
Babak
Schoeller
Anglin
Schoeller
Sandeep
Cho
Schoeller
Imran
Cotts
Arfaei
Francis
Cotts
Arfaei
Francis
Cotts
Arfaei
Anselm
Meilunas
Francis
Meilunas
Gaurang
Sa'D
Aaron
Cotts
Su
Borgesen
Su
Gaurang
Nick
Su
Park
Nick
Nick
Meilunas
Park
Kondos
Meilunas
Meilunas
Meilunas
Schoeller
Meilunas
Schoeller
Meilunas
Schoeller
Yang
Schoeller
Arfaei
Research Publications in 2014
A.R.E.A. Consortium High Reliability Session:
Material Evaluation for High Reliability Applications
• A.R.E.A. - Low Loss Laminate Material Pad Cratering Resistance
• A.R.E.A. - Effect of PCB Surface Finish on Sn Grain Morphology and
Thermal Fatigue Performance of SnPb and Lead Free Solder Joints
•
• A.R.E.A. - Component Level Testing of Thermal Interface Materials
• A.R.E.A. – Component Warpage: Issues with Measurement and
Standardization
Topic Expert Failure Analysis,
Method and Solutions
Denis Barbini, Ph.D.
Associate Director, Advanced Process Laboratory
603-828-2289
Barbini@uic.com
Tonight’s Focus
▫ Why are we asked to perform failure analysis?
− Third-Party Failure Analysis
▫ General Understanding of Root Cause Failure Analysis
▫ Case Studies
Why Perform Failure Analysis
▫ In a production environment, the appearance of failures is an
unfortunate inevitability.
▫ In every case Electronic Manufacturers take every precaution to reduce
the number of failures that occur in their facilities.
− This inherently breeds a lack of knowledge and understanding for the
evaluation of failures.
− Since failures are rare it is not financially justified to have a highly skilled
and trained dedicated workforce.
▫ To this end a laboratory that provides the service of failure analysis can
be positioned to have all the necessary equipment and recourses for
the determination of root cause.
▫ The Universal Advanced Process Laboratory takes this one step further
by having the advantage of a research organization as part of its
organization as well as a complete prototype manufacturing facility.
▫ Having experts in PCB fabrication, underfilling, rework, surface mount
assembly, wave soldering, stencil printing, encapsulation, reliability
testing, PCB design, etc. gives us a unique perspective on failure
analysis.
Why Outsource Failure Analysis
•Cost and Time
▫ Line down situations
▫ New product evaluations
▫ Quick turn analysis
•Lack of Analytical Techniques/Understanding
•Lack of Experience with Material and Process
•Non-Biased Evaluation of Issues
Failure Analysis
•Failure analysis of Electronic assemblies requires an
understanding of :
▫ PCB fabrication
▫ Underfilling
▫ Rework
▫ Surface mount assembly
▫ Wave soldering
▫ Stencil printing
▫ Encapsulation
▫ Reliability testing
▫ PCB design
▫ etc.
Failure Analysis Methods
•Depends upon the type of analysis being conducted
▫ Manufacturing failures require an understanding of the process,
material systems, components and the issues which drive production
related failures.
▫ Field failures require a detailed knowledge of the environmental
applied strains, material interactions and the possible failure modes.
Performing Failure Analysis
•Preparation is a vital part of proper analysis
▫ Can involve physical or chemical preparation
▫ Often destructive (careful not to affect the failure mode of interest)
▫ Often requires a significant amount of analysis time
− Conversations with customers
•Understanding of the failure conditions and the variables involved
▫
▫
▫
▫
Date codes/lots involved
Failure rate
Design and supplier changes
Failure description from customer
− Example: de-wetting and non-wetting
• What is the difference?
• Be careful most don’t know
• Non-wetting: Mask on pad, oxide, profile (More common failure mode)
• De-wetting: Black Pad
Thought process
Lead-free vs. Sn/Pb
• Circuit board issues are the biggest issue in LF and Sn/Pb (could be
related to high Tg of LF laminates)
▫ This is a big issue for the industry
▫ Brittle laminate materials lead to CAF, pad crater, etc.
• Other material issues such as BGA warpage and surface solderability
are affected by transition to LF
▫ Surface finish
▫ Look for Our Paper at SMTA International
• Manufacturing processes must change to accommodate known LF
materials and manufacturing issues
• If you don’t understand the issues then you will not be informed
enough to work with your suppliers to resolve problems and drive to
solution
Case Studies
• Random sampling of supplier related material issues we have
observed that do not have obvious root causes
• If you don’t know the root cause for the failure then;
▫ You can’t formulate an effecting corrective action plan
▫ You will not be in a position to demand new materials, or
reimbursement for lost revenue
▫ Risk lost time in manufacturing resulting in missed shipping dates,
product launches, increased WIP, etc.
• We don’t have much time so only a few topics are discussed
here
Case Study 2:
Component Functionality
QFN Failure
• Common issues with QFN’s
▫ Thermal pad voiding
▫ Process conditions
−Open or short conditions due to pad design and stencil design
−Tilted device due to solder volume variations
• Less common issue is related to clock speed variations
▫ Exhibits not a true shorting condition
▫ Caused by electrical leakage
• Chemical Analysis and sample prep
QFN Root Cause
• WS paste used for production, high halide content creates low
dielectric strength at T0 and then dendrite growth
QFN solution
• No clean solder paste should be used.
• Do not wash these devices
NB
• This was 4 years ago:
▫ Current cleaning chemistries and cleaners should be able to
easily wash and leave clean surfaces.
Case Study 3:
Aging Related Issues
Field Failures When Soldering to Cu-Pads
•
TI published work on voiding in Cu3Sn in
2004.
•
Many others have also seen it and report
it, but often not ‘on the record’.
•
Consequences & potential severity are
commonly underestimated.
•
•
We have established approach for
interpretation & extrapolation to service.
We can turn the problem on and off.
Case Study 4:
Shorting in the Board
BOARD RELATED ISSUES
• Board processing is complicated especially in HDI constructions.
• We have observed numerous failure analysis projects that are
related to board construction issues
• Failure mode determination can be simple
▫ Inter Connect Defect
▫ Plating
▫ Mask on pad
▫ Imbedded Foreign Material
• Root cause and fabrication improvements can be complicated and
often beyond the CM.
▫ What questions to ask your suppliers?
• The following slides will be related to board issues and will touch on
some of the questions that should be asked.
IFM – Imbedded Foreign Material
Traces severed at
these locations
X-ray image of shorted electrical network. Green arrows indicate open segment and red arrows indicate the shorted segment.
Back-lit illumination showing a potential conductive filament bridge, imbedded just below
the solder mask and highlighted in the image above.
Horizontal section showing location of the filament as seen from the bottom of the board, looking upward. Yellow arrows highlight the filament (foreign material).
Trace
Ground Plane
Flat section backlit and photographed at two focal depths showing intimate contact (red arrows) between filament, trace,
and ground plane. Filament was visually determined to be embedded between the solder mask and L1 dielectric.
Horizontal sectioning and evaluation determined that a thin filament of copper was embedded
between the core material (dielectric) and solder mask. Thus making intimate contact
between shorted trace and ground plane. A helical shape indicates that the filament was most
likely generated during the drill process and redeposited during subsequent processing.
Case Study 5:
Impact of Materials Due to
Pb Free Transition
Pad Cratering
•“Push Button” failures
•Pad cratering is often driven by external mechanical stress, however in order to predict whether the pad is
likely to fail we must consider the PCB pad design, substrate material, component design, etc.
•Factors that affect the preferred failure location are;
▫ pad diameter
▫ solder mask verses non-solder mask defined pad design
▫ trace width into the pad
▫ Location of the pad relative to the component,
▫ The weave
▫ and substrate material’s resistance to fracture
•These failure modes become more prevalent in lead-free assemblies due to the properties of the high Tg
laminate materials
▫ Common to hear “we were building with Sn/Pb for years and never had an issue, now that we are
assembling with lead-free materials…”
▫ Lead-free boards are being used in Sn/Pb military and medical products, higher probability of failure in
this mode.
Pad Crater – Case A
•Dye penetration identified a number of opens in
both the OSP and ENIG boards.
•For the ENIG board, three failure types were
observed: cracks at the component side, cracks at
the PCB side, and fractures under the PCB pads
(pad crater).
• For the OSP assembly the observations were
similar, but with only one component-side
fracture.
•The mixture of failure modes indicates the
problem was mechanical, with the assemblies
being subjected to high stresses.
Pad Crater – Case A
Pad craters observed in ENIG (left) and OSP (right) assemblies.
Pad Crater – Case B
Intermetallic fracture
Pad rupture
Pad Crater – Case C
Component Pad Rupture
Component intermetallic failure
Component Pad Rupture
u3110
u3120 pad rupture on the component pad and
through the intermetallic of the upper left joint
Component Pad Rupture
Reliability Issue
• Connecting traces, and/or Vias
will break, rendering device
non-functional.
• Damage is not easily re-worked,
so product is scrapped.
page 36
Strength Scaling
• Strength scales with pad
area, using both a
quadratic and linear term
Pad Strength (grams-force)
2500
2000
F
1500
1.56(d 2 ) 33.92(d )
1000
500
• Quadratic term is related
to pad area.
0
0
5
10
15
Pad Diameter (mils)
• Linear term is related to
crack depth.
37 June 22, 2011
20
25
30
Pad Cratering solutions?
• Board design
▫ Solder mask defined pads
▫ Placement of components
• Material selection
▫ Long list of materials tested within our research
consortium
• Adhesive
▫ Edge bond, corner bond, underfill
Case Study 6:
Poor Product Performance
BGA Voiding in HDI construction
• Voids due to PCB via-in-pad
fabrication issue
• Laser drilled holes
• Desmear was not
done correctly.
BGA Voiding in HDI construction
Voids due to PCB via-in-pad
fabrication issue.
•
Solutions are obvious but should
be discussed with board supplier
• Redesign should be considered
Case Study 7:
NPI Assembly Failure
Ceramic Capacitor Failures
• Fractures in devices
• Design of board must be scrutinized
▫ Pad size, mask thickness
▫ Proximity to edge of board
▫ Orientation
• Board warpage
• Final assembly handling
• Panel singulation
▫ Breakaways vs. router vs. pizza cutters
Capacitor failures
•Excellent White Papers
▫ KEMET; Ceramic Chip Capacitors “Flex Cracks” Understanding &
Solutions by Jim Bergenthal
▫ AVX; CRACKS: THE HIDDEN DEFECT by John Maxwell
•These failure modes have been well documented
▫ The issue is not with diagnosis
▫ The solution can be difficult and complicated
▫ Strain gauge analysis
Capacitor failures
• Example of a severe
failure on surface mount
capacitor
Pick and Place Failures
• Cap fractures
Lab Exercise
• What are the action items?
• _______________________
• _______________________
• _______________________
• _______________________
• _______________________
• _______________________
• _______________________
• _______________________
• _______________________
• _______________________
• _______________________
page 48
• _______________________
Lab Exercise
• Cracked capacitor
• Only images provided to you initially (top
down)
• Customer indicates that this is a unique
occurrence in prototype engineering
boards
• Wants to know if this will reoccur in
production
• One board is being provided since this is a 1
time occurrence
▫ What do you quote, what are the
action items once this board comes in
house?
page 49
Lab Exercise
• Consult white paper
documents like the MuRat
failure mode classification
▫ Kemet
▫ AVX
page 50
Lab Exercise
Optical inspection of product once received
• New ideas?
• Does this change your original
quote/action items?
Disturbed solder
page 51
Lab Exercise
• Lets look at the white papers again…
page 52
Lab Exercise
D1-D3
D4-D6
D7-D9
page 53
Lab Exercise
D3
D1
D4
D6
Disturbed solder
page 54
Case Study 9:
Component Failure
Wire Bond Failures of Pressure Sensor
Fracture initiation site
Area of high rate of
fracture propagation
Area of slow rate of
fracture propagation
Typical Laser Vibrometer Measurement Setup
• -An excitation (shaker, loudspeaker,
hammer etc.) causes the object under
investigation to vibrate.
• -The measurement beam from the
interferometer in the scanning head is
positioned to a scan point on the object by
means of mirrors and is scattered back.
• -The backscattered laser light interferes
with the reference beam in the scanning
head.
• -A photo detector records the interference.
• -A decoder in the controller provides a
voltage which is proportional to the velocity
of the vibration parallel to the
measurement beam.
• -The voltage is digitized and processed as
vibrometer signal.
(Theory Manual - Polytec Scanning Vibrometer)
Board / base displacement exported to Matlab
- Estimate the strain field from this data
For 1 g input, center displacement ~150 um
Summary Failure Analysis
▫ Failure Analysis can be used for production failures to
increase yields and improve product reliability.
▫ Field Failures can be effectively analyzed to determine
the cause of failure and aid in accurate and therefore cost
effective product recall.
▫ Failure Analysis can provide the evidence required to
support vendor returns in product liability cases.
▫ Failure analysis is a fast and cost effective method of
improving yields and product reliability.
▫ The ROI (return on investment) in Failure Analysis easily
justifies the minimal cost.