Root Cause Analysis of HALT Failures

© 2004 - 2007© 2004 - 20109000 Virginia Manor Rd Ste 290, Beltsville MD 20705 | 301-474-0607 | www.dfrsolutions.com© 2004 – 2010
Root-Cause Analysis of
HALT Failures
Cheryl Tulkoff and Greg Caswell
ctulkoff@dfrsolutions.com
gcaswell@dfrsolutions.com

© 2004 - 2007© 2004 - 20109000 Virginia Manor Rd Ste 290, Beltsville MD 20705 | 301-474-0607 | www.dfrsolutions.com
“the process of seamlessly
cohesively integrating reliability
tools together to maximize
reliability and at the lowest
possible cost”
Reliability Integration

Reliability and CostCOST
RELIABILITY
TOTAL
COST
CURVE
RELIABILITY
PROGRAM
COSTS
WARRANTY
COSTS
OPTIMUM
COST
POINT
Use of the proper tools during the proper life cycle phase will
help to minimize total Life Cycle Cost (LCC).

o To minimize total Life Cycle Costs (LCC), a Reliability
Engineer must do two things:
o choose the best tools from all of the tools available and must
apply these tools at the proper phases of a product life cycle.
o properly integrate these tools together to assure that the
proper information is fed forward and backwards at the
proper times.
o Highly Accelerated Life Testing (HALT) is one reliability
tool that can assist with this process
Reliability vs. Cost, continued

o Highly Accelerated Life Testing (HALT)
o In HALT, a product is introduced to progressively higher stress
levels in order to quickly uncover design weaknesses, thereby
increasing the operating margins of the product, translating to
higher reliability.
Introduction to HALT

HALT Testing Overview
o Typically exposes the product to simultaneous
vibration and thermal cycling
o Product is tested in operational mode while the
vibration stress is increased with each thermal
cycle
o Objective of the test is to cause failure of the
product
o Identifies the weakest link which can be then be
improved
o Test duration is typically less than a few weeks
o On its own, this test is not able to predict the life of a
product (acceleration factor is not known)
o However, it is very useful when a product can be
compared side-by-side with a previous generation of
product with known reliability

Example HALT Test
o Functional testing is performed while the vibration is taking place
o This is important since intermittent opens can be found at this condition
o Vibration starts at 0 Grms and step up by 5 Grms each cycle until failure is
detected
o Failing units shall be analyzed carefully to find root cause failure

o Uncovers flaws typically not found before product
introduction
o Discovers and improves design margins
o Reduces overall development time and cost
o Provides information for developing accelerated
manufacturing screens (HASS)
HALT - Advantages over Traditional Testing

HALT versus Traditional Reliability Testing: Comparison
HALT
o Gathers info on product
limitations
o Focus on Design Weaknesses
o 6 DoF (degrees of freedom)
Vibration
o High Thermal Rate of Change
o Loosely Defined - Modified
“On the Fly”
o Not a “Pass/Fail” Test
o Results used as basis for HASS
or ESS (environmental stress
screening)
Traditional
o Simulates Lifetime of use
o Focus on Finding Failures
o Single Axis Vibration
o Moderate Thermal Rate of
Change
o Narrowly Defined - Rigidly
Followed
o “Pass/Fail” Test
o Results typically not used in
ESS

HALT
Highly Accelerated
Life Testing

o Value of HALT
o Rapidly discover design issues.
o Evaluate & improve design margins.
o Release mature product at market introduction.
o Reduce development time & cost.
o Evaluate cost reductions made to product.
o Developmental HALT is not really a test you pass or
fail, it is a process tool for the design engineers.
o There are no pre-established limits.
HALT - Highly Accelerated Life Test

HALT, How It Works
Start low and step up the
stress, testing the product
during the stressing

HALT, How It Works
Gradually increase
stress level until a
failure occurs

HALT, How It Works
Analyze
the failure

HALT, How It Works
Make
temporary
improvements

HALT, How It Works
Increase
stress and
start
process
over

HALT, How It Works

HALT, How It Works
Fundamental
Technological
Limit

Product
Operational
Specs
Stress
Upper
Oper.
Limit
Upper
Destruct
Limit
Lower
Destruct
Limit
Lower
Oper.
Limit
Margin Improvement Process

Product
Operational
Specs
Stress
Upper
Oper.
Limit
Upper
Destruct
Limit
Lower
Destruct
Limit
Lower
Oper.
Limit
Margin Improvement Process
Operating
Margin
Destruct
Margin

o Combined stresses to technology limits
o Step stressing (individual and combined)
o Powered product with monitored tests
o Root cause failure analysis and appropriate corrective
action
HALT Implementation Requirements

Commonly Used HALT Equipment
o Combined
Temperature/Vibration
Equipment
o Pneumatic Vibration (to
provide the random
vibration) with Wide
Frequency Spectrum
o Fast Thermal Rates of Change
and Wide Thermal Range

o Planning a HALT
o Setting up for a HALT
o Executing a HALT
o Post Testing
Developing a HALT Process

o Meet with design engineers to discuss product.
o Determine stresses to apply.
o Determine number of samples available.
o Determine functional tests to run during Dev. HALT.
o Determine what parameters to monitor.
o Determine what constitutes a failure.
o Develop Test Plan
Planning a HALT
It is essential that the product being tested be fully exercised
and monitored throughout HALT for problem detection.

For each stress, we use the
Step Stress Approach
Continue until operating & destruct limits
of UUT (unit under test) are found or until test equipment
limits are reached.
Stimuli
Time (hour:minute)
D
C
B
A
0:00 0:10 0:20 0:30
Planning a HALT (cont.)

OTHER STIMULI:
• Voltage/frequency margining
• Power cycling
• Combined environment (Temp/Vib)
• Rapid transitions up to 60oC/min on the product
VIBRATION HIGH TEMP LOW TEMP
START
INCREMENT
DWELL TIME
END
3-5 G’s
3-5 G’s
10 min* 10 min* 10 min*
5 to 10° C
+20° C +20° C
Destruct Limit or Test Equipment Limitation
STIMULI
5 to 10° C
* In addition to functional test time
Planning a HALT (cont.)

Examples of Failure Inducing Loads
• Temperature Cycling
– Tmax, Tmin, dwell, ramp times
• Sustained Temperature
– T and exposure time
• Humidity
– Controlled, condensation
• Corrosion
– Salt, corrosive gases (Cl2, etc.)
• Power cycling
– Duty cycles, power dissipation
• Electrical Loads
– Voltage, current, current density
– Static and transient
• Electrical Noise
• Mechanical Bending (Static and Cyclic)
– Board-level strain
• Random Vibration
– PSD, exposure time, kurtosis
• Harmonic Vibration
– G and frequency
• Mechanical shock
– G, wave form, # of events

o Design vibration fixture to ensure energy transmission to the
product (different from electrodynamic vibration fixtures).
o Design air ducting to ensure maximum thermal transitions on
the product.
o Tune chamber for product to be tested.
o Apply thermocouples to product to be tested.
o Setup all functional test equipment and cabling.
Setting up for HALT

o Begin with cold step stress and then hot step stress.
o Step in 10 °C increments, as approach “limits” reduce to 5 °C.
o Dwell time minimum of 10 minutes + time to run functional tests to
ensure product is still functional. Start dwell once product reaches
temperature setpoint, begin functional tests after 10 minute dwell.
o Continue until fundamental limit of technology is reached.
o If circuits have thermal safeties, ensure operation & then defeat to
determine actual operating & destruct limits
o Apply additional product stresses during process:
o Power Supplies: Power cycling during cold step stress.
o Input voltage variation. Load variations.
o Frequency variation of clocks.
Executing a HALT (Thermal Step Stress)

o Transition temperature as fast as chamber will allow.
o Select temperature range within 5° of the operating limits
found during thermal step stress.
o If product cannot withstand maximum thermal transitions,
decrease transition rate by 10 °C per minute until
operating limit is found.
o Continue series of transitions for a minimum of 10 minutes
(or time it takes to run set of functional tests).
o Apply additional product stresses during process
Executing a HALT (Fast Thermal Transitions)

o Understand vibration response of product (i.e. how does
product respond to increases in vibration input).
o Determine Grms increments (usually 3-5 Grms on product).
o Dwell time minimum of 10 minutes + time to run functional
tests to ensure product is still functional. Start dwell once
product reaches vibration setpoint.
o Continue until reach fundamental limit of technology.
o Apply additional product stresses during process.
Executing a HALT (Vibration Step Stress)

Power Spectral Density (measured on product on OVS-2.5HP)
Marker Fcn
[Band Pwr]
Trace: D Strt: 0 Hz
Stop: 3 kHz
Band:5.468 grms
32Hz 12.8kHzAVG: 20
X:288 Hz Y:2.82893 m*
Band:6.548 grms
Band:9.842 grms
X:288 Hz Y:176.027 u*
32Hz 12.8kHzAVG: 20
X:288 Hz Y:3.44536 m*
32Hz 12.8kHzAVG: 20
Band:4.488 grms
A: ROBOT BD
B: ROBOT ARM
X:288 Hz Y:1.19064 m*
Y* = grms^2/Hz100
m*
10
u*
LogMag
Y* = grms^2/Hz100
m*
1
u*
LogMag
C: R MOTOR
D: TABLE
Y* = grms^2/Hz1
*
10
u*
LogMag
Y* = grms^2/Hz1
*
100
n*
LogMag
32Hz 12.8kHzAVG: 20
Date: 03-14-97 Time: 09:46:00 AM
Y-AXIS
Z-AXIS
X-AXIS
Z-AXIS
Vibration Step Stress (cont.)

Power Spectral Density (measured on product on QRS410T)
Vibration Step Stress (cont.)

o Develop thermal profile using thermal operating limits,
dwell times and transitions rates used during thermal step
stress & fast thermal transitions.
o Incorporate additional product stresses into profile such as
power cycling.
o The first run through the profile, run a constant vibration
level of approx. 5 Grms. Step in same increments
determined during vibration step stress.
o When reach higher Grms levels (approx. 20 Grms) add
tickle vibration (approx. 5 Grms) to determine if failures
were precipitated at high G level but only detectable at
lower G level.
Executing a HALT (Combined Environment)

o Determine root cause of all failures that occurred.
o Meet with design engineers to discuss results of
Developmental HALT and root cause analysis.
o Determine and implement corrective action.
o Perform Verification HALT to ensure problems fixed and
new problems not introduced.
o Periodically evaluate product as it is subjected to
engineering changes.
Post Testing

o Highly Accelerated Life Testing (HALT) is designed to
induce product failure
o Requires effective root-cause analysis
o Determination of failure site, failure mechanism, and root-
cause
o Absence prevents improvement in design margins; greatly
reduces value of HALT process
o DfR Solutions takes a systematic approach to root-
cause analysis
o Proceeding from the least destructive to most destructive until
root-cause(s) are conclusively identified
o Process steps based upon failure information (failure mode,
failure site, failure mechanism)
Failure Analysis (Motivation)
41

© 2004 - 2007© 2004 - 20109000 Virginia Manor Rd Ste 290, Beltsville MD 20705 | 301-474-0607 | www.dfrsolutions.com 42
No Analysis? No Improvement!

© 2004 - 2007© 2004 - 20109000 Virginia Manor Rd Ste 290, Beltsville MD 20705 | 301-474-0607 | www.dfrsolutions.com43
Information Gathering
o Crucial first step in any failure analysis effort
o Use of standard nomenclature
o Failure history
o When in HALT did the failure occur?
o Step stress testing, temperature cycling, vibration, combined
o What were the stresses being experienced by the product
before failure?
o Temperature, vibration, combined
o Also electrical stresses
o Is there a specific change in design/bill of
materials/manufacturing that is being assessed?
o Failure mode – the failure behavior experienced by the
observer
o Failure site – the supposed location of the failure
o Failure mechanism – the mechanism initiating the failure (not
root cause)

Failure History
o The stresses experienced before failure provides
strong indication of potential failure site
o Temperature step stress testing tends to drive
component issues
o High temperatures
o Driven by increase in leakage current (diode turn on) or increase
in resistance (time delay)
o Semiconductors, power components
o Cold temperatures
o Primarily liquid electrolytic capacitors
o Cyclic stresses (temperature cycling, vibration) tend to
drive interconnect issues
o Separable (sockets, press-fit connections, etc.)
o Permanent (solder joints)

Non-Destructive Evaluation (NDE)
o Designed to provide maximum information with minimal
risk of damaging or destroying physical evidence
o Evaluation methods
o Electrical Characterization
o Visual Inspection
o Xray Microscopy
o Thermal Imaging
o SQUID Microscopy
o Acoustic Microscopy

o Most critical step in failure analysis (FA) process for
product subjected to HALT
o Especially for failures induced during temperature step stress
testing
o Recoverable failures (operational limits)
o Failure site can only determined through electrical
characterization and interpretation of changes in functional
behavior
o Permanent failures (destruct limits)
o Proceeds with a broader range of possible electrical
characterization techniques
Electrical Characterization
46

o JTAG (joint task action group) Boundary Scan
o Very accurate identification of failure site; rarely performed on failed units
o Oscilloscope
o Useful in probing operational circuitry
o Resistance measurements (manual)
o Binary approach (best for an electrical open)
o Isolation of attached components
o Attempt to perform electrical characterization without component removal
o Parametric characterization
o Comparison of performance to datasheet specifications
o Curve tracer
o Valuable in characterizing diode, transistor, and resistance behavior
o Time domain reflectrometry (TDR)
o Measurement of phase shift of return signal can indicate location of electrical open
o Other characterization equipment
o LCR Meter, High resistance meter, low resistance meter, etc.
o Sometimes requires re-introduction of environmental stresses
o Especially elevated temperatures
Electrical Characterization Techniques
47

o Latest innovations
o Digital detector, laminography, nanofocus
resolution, oblique viewing
o Greatest potential
o Solder joint failures under area array
devices
o Problem
o Difficult to identify microcracks that occur
during fatigue/fracture of the solder
o Can detect voiding or poor wetting
o Provides evidence of root-cause of failures
during the application of cyclic stresses
Xray and HALT
48

o Best for component failures
o Especially increases in elevated leakage
currents
o Rarely used
o Electrical characterization often provides
more rapid identification of failure site and
more definitive information on failure
mode
o However, can provide evidence of
excessive temperature rise due to power
dissipation
o Can identify root-cause of failures at
temperatures below expectations
Thermal Imaging and HALT
49

SQUID Microscopy
 Current flow in devices produce a magnetic
field
 SQUID uses a highly sensitive magnetic detector
(superconductor) to resolve these fields
 Magnetic field image is converted to a current
density image, allowing for fault location
 Resolution
 500 nA, 300 nm
 Dependent on working distance (requires a flat
sample)

SQUID Microscopy
 Critical technology for
detecting package level
electrical shorts
 Much more rapid failure
site resolution
 Absolute confirmation of
shorting path
 Thermal imaging induces
damage

Failure Analysis: Temperature Tools
o Cold Spray and hot
plates to simulate fails
at temperature
extremes

Failure Analysis Dremel Tool – Induce Vibrations
o A Dremel tool can be
used to induce local
vibration during
debugging
o http://www.dremel.com

Sherlock Analysis
o Sherlock Automated Design Analysis™ can be used to assess the validity of
failure mechanism
o Can provide prediction of modes shapes and stresses during vibration
o Compare prediction to observation
o Were vibration loads sufficient to fail the interconnect?
o Were defects identified during failure analysis?
o If answers to one and two are no, maybe issue with test procedure
o Important: Can not be used for combined environment (temperature cycling
+ vibration)
o Too complex
340 Hz (369 Hz) 806 Hz, (715 Hz) 945 Hz, (910 Hz)

NDE and HALT
o Failure site identification is best performed through
electrical characterization, visual inspection, and an
understanding of failure mechanisms
o NDE techniques are most useful in the second phase of
failure analysis, when root-cause must be determined

Root-Cause Evaluation
o Appropriate point to evaluate information obtained
o Requires understanding of failure mechanisms and root-cause
o Inappropriate technique could destroy evidence
o Use of failure analysis tools to direct destructive phase
of process
o Fault tree analysis (FTA)
o Fishbone diagram
o Stress vs. Strength
o Failure modes and effects analysis (FMEA)?
o Useful for design review; rarely used in failure analysis
process

o Used once approximate failure site has been
determined
FTA (cont.)
57
Ceramic
Capacitor
Failure
Insufficient
Capacitance
Dielectric
Formulation
Crack
Propagation
Solder Joint
Open
Elevated
Leakage
Current
Crack
Propagation
Surface
Insulation
Resistance

Fishbone Diagram (Ishikawa)
o Useful in identifying all potential influences that could
drive product failure
o Can result in overcommitment of resources

Stress vs. Strength
59
 Fiber/resin separation
 Copper/resin separation
 Hollow fibers
 Misregistration of PTHs/Vias
 Copper wicking
 Drilling damage
 Separation of PTH wall from epoxy resin
 Higher voltage
 Higher moisture concentration
 Tighter conductor spacing
 Higher ion concentration in epoxy resin

Destructive Evaluation
o Used when information obtained during NDE is
insufficient to conclusively identify root-cause of failure
o Primarily consists of two techniques
o Cross-sectioning
o Decapsulation/delidding

o Critical for interconnect failures (cyclic stresses)
Cross-Sectioning and HALT
61

Cross-Sectioning
o Identify the area of interest (failure site)
o Remove the area of interest
o Diamond-encrusted blades are preferred (minimizes surface damage)
o Band saw, high and low speed diamond saw
o Plunge cuts preferred (minimizes damage to surrounding PCBA)
o Mount sample
o Prevents relative movement / improves handling
o Use room-temperature cured epoxy
o Plane of interest should be near the surface of sample
o Excessive removal of material increases likelihood of non-planarity,
overheating, and loss of root-cause information

Grinding/Polishing
o Rough grinding
o Used to approach the area of interest
o Involves application of 60 to 400 grit
of SiC paper
o Rough polishing
o Used to remove damage introduced
during grinding
o Involves 600 to 1200 grit of SiC paper
o Fine polishing/etching
o Used to reveal microstructural elements
o Involves use of polishing media
one micron and finer
Grit
(Silicon Carbide)
Particle Size
(microns)
240 50
320 29
400 17
600 9
800 7
1200 3

o Use of acid to remove encapsulant and
expose internal workings of the
package
o Used to analyze on-die failures
o Primarily driven by hot step stress test
o Interconnect issues, primarily wirebonds,
that can be discovered during
temperature cycling, are best initially
investigated using cross-sectioning
o Further characterization, using mechanical
testing, will require decapsulation
Decapsulation and HALT
64

Case Study -- Cold Step Stress Test
o Mass flow meter
o Recoverable failure at -30C
o Failure mode
o Loss of communication
o No permanent failures observed
o Results of electrical characterization / functional testing
o Insufficient filtering of electrolytic capacitors (rated at 105C)
o Parametric testing identified drop in capacitance at -35C
o Freezing of the electrolyte
o Corrective actions that were considered
o Switch from liquid electrolytic capacitor to tantalum capacitor
o Switch from 105C rated to 85C rated (reduced lifetime)
o Increase capacitance from 3.3 uF to 47 uF
o Extends range as well as improves filtering

Cold Step Stress Test (cont.)
o Subsequent testing on production units identified an additional
failure mode
o Initiated at approximately -5C
o Out of specification on line voltages, corrupting micros in the unit
o Incorrect resistor values
o Resistors were removed and replaced with different values
o Units passed remaining cold step stress test
o Of interest, the unit used for the initial HALT did not demonstrate
these cold temperature problems, showing impact of component
variability on product performance
o Identifying design margins!
o Subject sufficient number of samples to HALT process

Case Study -- Hot Step Stress Test
o Mass flow meter
o Permanent failure at 140C
o Failure site
o Catch diode for a switching power supply
o Failure mechanisms
o Electrical short (< 1 ohm). Operating junction temp for that
part is -65C to 125C.
o Diode was replaced and the unit was functional after
exposure at 140C. Temperature was stepped up to
150C, where nonfunctional failure reoccurred

Hot Step Stress Test (Example)
Hello Craig,
We found that an Optical-Isolator started getting noisy at higher temperatures.
The communication link stopped updating at 50C because digital data (1's and
0's) were getting through, but at a reduced amplitude. Thus, some of the 1's
were being misinterpreted as 0's by the foundation fieldbus communication
card.
We solved the problem by getting a better Opto-isolator with a higher current
gain.
Best regards,
Jim

o Application of vibration, starting at 5g and increasing
in 5g increments
o First failure noted at 30g
o Failure site identified as connector solder joints
o Insufficient flow through
o After touch up unit survived up to 40g
o Incorrect approach to failure analysis
o Unit was fixed as soon as a problem was detected
o Root-cause unable to be identified
o Design for reliability? Design for manufacturing? Processing
defect?
Case Study -- Vibration
69

o Failure after exposure to vibration
o Electrical characterization indicated electrical open under
area array device
o Confirmed through cross-sectioning
Case Study – Desktop Computer
70

o Validity of failure mechanism?
o Additional observation of
separated leads on electrolytic
capacitors
o Occurrence of failure mechanism
was strongly dependent upon the
orientation of the leads
o Lead planes oriented along the
board length most likely to fail
o Vibration test may not have
applied random loads
o Potential issues with HALT table or
fixture
Issues with HALT
71

When FA and HALT?
o Failure analysis can be a time intensive process
o Hold up in product release while awaiting results
o The use of FA should be selective and should provide
maximum value

FA and HALT (cont.)
o Product
o When product design or functionality is revolutionary, perform FA on
all failures
o When product design or functionality is evolutionary, perform FA
selectively
o Temperature Step Stress Test
o When recoverable failure occurs between the operational and
storage specifications
o Specified to operate between 0 to 70C
o Specified for storage between -40 to 100C
o E.g., recoverable failure occurs at 90C
o When permanent failure occurs within 10C of cold temperature
storage specification or within 20C of hot temperature storage
specification

FA and HALT (Cyclic Stresses)
o Delineation between when to perform FA less definitive
o General rule
o Temperature cycling should not induce any failures, unless
using custom designed interconnect
o Use prior behavior to guide failure analysis in vibration or
combined
o Failure on previous designs is always the electrolytic capacitor at
20g’s
o Identification of processing defect can be a design
issue!
o Design for manufacturing

Conclusion
o HALT can be an important step in best practice
reliability activities
o Use can be extremely limited without root-cause
analysis
o Value-added root-cause analysis requires
understanding of failure mechanisms and the stresses
that drive them
o Sufficient knowledge base allows for optimization of
resources and rapid feedback

Root Cause Analysis of HALT Failures

More Related Content

What's hot (6)

Similar to Root Cause Analysis of HALT Failures (20)

Recently uploaded (20)

Root Cause Analysis of HALT Failures