a printer-friendly PDF version

Frequently Asked Questions About
PEM Reliability
Orlando, Florida
February 6 and 7, 1998
Updated October 3, 2000
W.L. Schultz
S. Gottesfeld
PEM Cpmsprtium 00/mmg
Subject Outline
ÿ
Plastic (PEMs) vs. Hermetic (HSMs)
ÿ
Moisture Sensitivity/Al Corrosion
ÿ
Moisture Sensitivity/SMD Popcorning
ÿ
Temperature Cycle Stress
ÿ
Plastic Molding Compound
ÿ
PEM Temperature Range
ÿ
Manufacturing Controls for Reliability
ÿ
Conclusions
PEM Cpmsprtium 00/mmg
Plastic (PEMs) vs. Hermetic (HSMs)
PEM Cpmsprtium 00/mmg
Plastic (PEMs) Vs. Hermetic (HSMs)
Are PEMs as reliable as HSMs?
ÿ PEMs will provide reliability equivalent to or better than HSMs
if:
– Application environment is understood relative to PEM
capability.
– Industry’s best practices are used by supplier and user.
– Sufficient PEM data/experience exists for similar application
environments
PEM Cpmsprtium 00/mmg
Plastic (PEMs) Vs. Hermetic (HSMs)
How does PEM failure rates compare to that of HSMs?
ÿ Intersil data for same technologies and device types have shown comparable
failure rates on High Temperature Operating Life tests conducted in same
time frame.
Failure Rate at 55°C, 60% UCL
Package
Sample (1)
FITs (2)
Hermetic (CERDIP
117,991
7.9
Plastic (PDIP)
342,252
8.2
1. Combines data (1987-94) on CMOS and Bipolar IC’s in 8 to 20-Ld packages.
2. Extrapolated from stress temperatures (³ 125C), FITs = Fails in 109 Device
Hours, UCL = Upper Confidence Limit.
PEM Cpmsprtium 00/mmg
Plastic (PEMs) Vs. Hermetic (HSMs)
What are the differences in failure modes and mechanisms?
ÿ HSMs have as many failure modes and mechanisms as PEMs.
ÿ HSMs more susceptible to mechanical stress.
ÿ Molded PEM construction immune to mechanical stress.
ÿ PEMs more susceptible to thermal and moisture stress.
ÿ Comparative listing of modes/mechanisms shown in following
chart.
PEM Cpmsprtium 00/mmg
Package Related Failure Modes/Mechanisms
Description
Stress/
Source
Cracked Die
Thermal
Electrical Short/Open Temperature Cycle
Mechanical
Electrical Short/Open Impact Shock
Thermal
Electrical Open
Temperature Cycle
Mechanical
Electrical Open
Vibration, Centrifuge
Thermal
Electrical Open
Temperature Cycle
Mechanical
Electrical Open
Vibration, Centrifuge
Thermal
Electrical Open
High Temp Storage
Thermal
Loss of Hermeticity
Temperature Cycle
X
Mechanical
Loss of Hermeticity
Impact Shock
X
Moisture
Loss of Hermeticity
Humidity,
X
Wire Breaks
Wire Lifts
Wire Lifts
Response
Accelerating
Test
Plastic Hermetic
X
X
X
X
X
X
X
X
X
X
X
(Intermetallic)
Cracked Seals, external
Corroded Seals, external
(Pin-to-Pin Shorts)
Interface Delamination
Salt Atmosphere
Thermal
Reduced Moisture
Temperature Cycle
X
PEM Cpmsprtium 00/mmg
Package Related Failure Modes/Mechanisms
(Continued)
Description
Stress/
Source
Internal Water Vapor Package
Response
Al Corrosion
Assembly
Moisture Ingress
Moisture
Accelerating
Test
Plastic Hermetic
Low Temperature
X
Bias Life
Al Corrosion
Temp/Humidity/Bias
X
Humidity/Solder Shock
X
Autoclave, HAST
SMD Cracked Pkg.
Thermal
(Popcorn Effect)
Metal Deformation
Reduced Moisture
Resistance/Elect. Opens Sequence
Thermal
Electrical Shorts/Open
Temperature Cycle
Thermal
Electrical Shorts/Open
Temperature Cycle
X
Cracked Passivation
Lifted Die
Mechanical Thermal Designation
Impact Shock, Centrifuge
Die Attach Voids
Package
Assembly
Thermal Dissipation
Low D/A Strength
Cracked Die
Bias Life
Temp Cycle,
Centrifuge
Loose Die Attach
Package
Electrical Shorts
Vibration/Shock PIND
X
X
X
X
PEM Cpmsprtium 00/mmg
Moisture Sensitivity/Al Corrosion
PEM Cpmsprtium 00/mmg
Moisture Sensitivity/Al Corrosion
Since PEMs are nonhermetic, why don’t all parts corrode?
Presence of moisture alone is not likely to cause corrosion.
Corrosion is chemical reaction which requires the presence of an ionic species,
such as chloride, which sets up an electrolytic cell with moisture.
Rate of corrosion is function of combined effect of :
– Bias Voltage
– Moisture
– Temperature
– Conductivity of Electrolyte
PEM Cpmsprtium 00/mmg
Moisture Sensitivity/Al Corrosion
How does ionic contamination cause corrosion?
Chloride is the most common ionic contaminant inducing corrosion.
– Al oxide first dissolved by Cl ion: Al(OH)3 + Cl- → AL(OH)2Cl + OH– Exposed Al reacts with Cl ion: Al + 4Cl - → [AlCl4]- + 3e– Al anion reacts with water: 2[AlCl4] + 6H2O→ 2Al(OH)3 + 6H+ + 8Cl– The corrosion product is aluminum hydroxide.
Reaction frees up the Cl ion to continue the process as long as moisture is present.
PEM Cpmsprtium 00/mmg
Path of Ingress of Ions and Moisture
PEM Cpmsprtium 00/mmg
Moisture Sensitivity/Al Corrosion
Is moisture induced corrosion a concern today?
Main deterrent in use of PEMs by Military dating to 1970’s.
Survivability has significantly increased in last 15 years:
– Improved mold compound purity and adherence
– Improved purity of die attach materials
– Improved glass passivation and deposition tools
– Improved leadframe construction and design
– Cleaner and more automated manufacturing processes
– Elimination of halide fluxes and other sources of halides
– Education of the user relative to board assembly of PEMs
Corrosion is rarely the cause of failure today (PEMs used in billions/year).
PEM Cpmsprtium 00/mmg
Al Corrosion Wearout Characterization at 85% RH Demonstrates Up To A 6X
Increase in MTF Due to Improvements In Mold Compound, Lead Frame
Construction, and Cleaner Processing
PEM Cpmsprtium 00/mmg
Comparison of Plasma Enhanced Nitride (PEN) and Phosphosilicate Glass
(PSG) Passivation for Al Corrosion
PEM Cpmsprtium 00/mmg
THB 85C/85% RH, CMOS Logic IC’s in PDIP
THB 85C/85% RH, CMOS Logic IC's in PDIP
Cumulative % Failure at 1000 Hours
100
10
1
0.1
0.01
1974
1979
1984
1989
1994
1997
Year
PEM Cpmsprtium 00/mmg
Moisture Sensitivity/Al Corrosion
What acceleration factor models are used for corrosion?
Intersil wearout data showed good fit to Peck’s model.
AF = exp [Ea/k (1/TU - 1/TS)] (RHS/RHU)a (VS/VU)b
where
AF = Acceleration Factor
T = Temperature at Use and Stress (°C + 273)
RH = Relative Humidity at Use and Stress Conditions
V = Voltage at Use and Stress Conditions
Ea = Activation Energy (0.9 eV) from Intersil Data
k = Boltzman’s Constant (8.62 E-5 eV/°K)
a = 2.66 Based on Peck [1]
b = 1.4 from Intersil Data
[1] Peck and Hallberg, Quality and Reliability Engineering International, (1991).
PEM Cpmsprtium 00/mmg
CMOS Logic Plastic Dual-In-Line Package (PDIP) Life Prediction To
1% Failure For Aluminum Corrosion As A Function Of Temperature
and % Relative Humidity at 5 Volts Continuous Bias
PEM Cpmsprtium 00/mmg
Moisture Sensitivity/Al Corrosion
What are the results of moisture-related tests used to monitor PEMs?
Data obtained during 1997 on 14 - 28 lead PDIPs and SOICs:
Test & Conditions
Temp/Humidity/Bias
85C/85% RH/Rated VDD
HAST
135°C/85% RH/Rated VDD
Autoclave
121°C/100% RH/15 psig
Failures/Samples
Hours
PDIP SOIC
1000
0/3858
0/1459
48
0/4410
0/2385
96
192
0/7290
0/9122
0/2385
0/7199
PEM Cpmsprtium 00/mmg
Long Term 85°°C/85% RH Test Data
Test
Storage
(No bias)
Temperature/Humidity/Bias
(6V bias)
Temperature/Humidity/Bias
(18V bias)
*
Package
Hours
PDIP
PDIP
PDIP
PDIP
SO*
PDIP
SO*
PDIP
PDIP
PDIP
PDIP
PDIP
PDIP
PDIP
PDIP
7,000
10,000
11,000
3,000
3,000
5,000
5,000
7,000
13,000
14,000
17,000
5,000
7,200
8,000
9,000
Sample
40
120
40
180
290
50
50
40
40
40
40
40
40
120
160
Failures
0
0
0
0
0
0
0
0
0
0
0
0
1 @ 7.2k
0
1 @ 6k
1 @ 9k
Data combines Digital Logic (CMOS Metal Gate, CMOS Silicon Gate) and Bipolar ICs.
Combines parts subjected to various conditions of reflow solder pre-conditioning.
PEM Cpmsprtium 00/mmg
Moisture Sensitivity/SMD Popcorning
PEM Cpmsprtium 00/mmg
Moisture Sensitivity/SMD Popcorning
What is “popcorn cracking”?
SMDs subjected to solder reflow temperatures are susceptible to cracking and
delamination of mold compound.
Temperatures create sudden vaporization of absorbed moisture
– Absorbed moisture > 0.2% of package weight critical
– Larger/thinner packages more susceptible
Cracking creates paths for ingress of moisture/contaminants.
Phenomenon is industry wide materials problem.
Current solution is dry-packing and floor life control of sensitive SMDs.
PEM Cpmsprtium 00/mmg
Popcorn Cracking
PEM Cpmsprtium 00/mmg
SEM Photos of Popcorn Cracking
MQFP Package
Die
Mold Compound
Mold Compound
Crack
Lead Frame
Paddle
Crack
Crack on External Package Surface
Cross-Section of Crack Propagation
PEM Cpmsprtium 00/mmg
Fick’s Law of Diffusion
1.
Mt1
=4
MS
2.
3.
1n
(
MS - Mt2
M(x, t)
Mtl, Mt2
MS
1
package
MS
(
)
= MS
Df
tl
p
12
= 1n
½
)
( )
8
p2
( å(
4
p
2
- p
Df t22
12
1
exp
2k+1
(
= water absorption ratio at time t1 and t2
= saturated moisture absorption ratio
= thickness of
- (2k+1) 2 p
12
Df
x
t
2
Df t)
) (
sin
)))
(2k+1) p 2 x
1
= moisture diffusion coefficient
= distance from package edge to paddle surface
= moisture absorption time
PEM Cpmsprtium 00/mmg
PLCC 68-Lead Moisture Absorption/Desorption
PEM Cpmsprtium 00/mmg
Different Mold Compounds Moisture Absorption 45°°C/85% RH
PEM Cpmsprtium 00/mmg
Moisture Sensitivity/SMD Popcorning
How is SMD moisture sensitivity determined?
JESD22 - A112/A113 specifies standards for pre-conditioning, characterizing and
rating moisture sensitivity levels of SMDs.
Level
JESD22 - A112 Moisture Sensitivity Levels
Dry Pack Req’d
Floor Life*
1
No
Unlimited @£30°C/90% RH
2
Yes
1 Year
3
Yes
168 Hours @ £ 30°C/60%RH
4
Yes
72 Hours @ £ 30°C/60%RH
5
Yes
24 Hours @ £ 30°C/60%RH
6
Yes
6 Hours
@ £ 30°C/60%RH
@ £ 30°C/60%RH
* Conditions apply after removal from dry pack container.
PEM Cpmsprtium 00/mmg
Temperature Cycle Stress
PEM Cpmsprtium 00/mmg
Temperature Cycle Stress
What effect does temperature cycle have on PEMs?
Temperature cycle works the CTE (coefficient of thermal expansion) of each
material in contact with another.
PEM Material
CTE (ppm/°°C)
Chip Silicon
Chip Oxide
Chip Aluminum
Leadframe (Alloy 42)
Leadframe (Copper)
Gold Wire
Die Attach Material
Mold Compounds
3.5
8 - 10
23.8
5.0
16.9
14.3
40 - 70
7.5 - 28
PEM Cpmsprtium 00/mmg
Temperature Cycle Stress
What effect does temperature cycle have on PEMs ? (continued …)
There should be no detrimental effects within the specified temperature cycle range.
Under excessive temperature cycle stress, differences in material CTE can lead to:
– Delamination of mold compound to internal surfaces (die, leadframe)
– Cracked passivation and interlevel oxide at die corners (large die)
– Deformation of chip metallization at die corners (large die)
– Bond wire lifts/breaks at die corners (large die)
PEM Cpmsprtium 00/mmg
Temperature Cycle Stress
What practices are used to prevent temperature cycle failure
mechanisms?
Best practices involve:
– Lower stress mold compounds (larger die/packages)
– Design rules for metal layout and bond placement
– Passivation stress characterization
– Planarized die surfaces
– Adherence to temperature cycle ratings by user
– Stress-relief die coatings (ultra sensitive devices)
PEM Cpmsprtium 00/mmg
Temperature Cycle Stress
What are the temperature cycle acceleration factors?
Stress is modeled by the Coffin-Manson relationship:
AF = (∆
∆TS /∆
∆TU) n
where,
AF = Acceleration Factor
∆T = Temperature excursion at stress and use
n = Exponent based on failure mechanism
Values of n reported: n = 4 for bond wire fatigue
n = 7 for bond/silicon fracture
n = 11 for thin film cracking
PEM Cpmsprtium 00/mmg
Modeling Stress: Stress Maximum at Chip Corner
PEM Cpmsprtium 00/mmg
Power Slug SOIC -65°C to +150°C Temperature Cycle
# Lifts < 3g
Post Stress Wire Pull @ 1000 Cycles
Lifts <3g by Pad Location (2 Mil Au Wire)
ÿ
ÿ
Pin 16
Pin 1
Bond Pad
Pin 23A
PEM Cpmsprtium 00/mmg
Power Slug SOIC
Temperature Cycle -65°°C to +150°°C
Delamination Progression at Die Corners
(From C-SAM Analysis)
After 1k cycles
After 2k Cycles
Die
Bond Pad
PEM Cpmsprtium 00/mmg
Power Slug SOIC Temperature Cycle Wearout For Corner Bonds
PEM Cpmsprtium 00/mmg
Power Slug SOIC Temperature Cycle Wearout For Corner Bonds
Condition
DT (°C)
t50 (Cycles)
Sigma (s)
-40°C to +125°C
165
2700
.378
-40°C to +150°C
190
1200
.344
-65°C to +150°C
215
840
.405
1) Log-normal median-time-to-failure (t50) is computed for corner bond
lifts <3.5.
2) Average of the log-normal s values = .375.
Power Law Regression Line Fit:
t50 = 1.8 (DT) 4.4, N = 4.4, R2 = 0.97
Power Law Acceleration Factor Model:
AF = t50 (T) / t50 (U) = (DTT / DTU) N
Where, AF =
DTT =
TU =
N =
acceleration factor
temperature excursion at test conditions
temperature excursion at use conditions
4.4 (from data)
PEM Cpmsprtium 00/mmg
Number of Temperature Cycles Equivalent to
Qualification Stress Conditions
PEM Cpmsprtium 00/mmg
Plastic Molding Compound
PEM Cpmsprtium 00/mmg
Plastic Molding Compound
What is mold compound glass transition temperature (Tg)?
Tg is the temperature corresponding to the glass-to-liquid transition:
–
Below Tg: CTE is relatively low and increases slightly
over temp.
–
Above Tg: CTE is high and increases substantially
over temp.
PEM Cpmsprtium 00/mmg
Plastic Expansion Curve
PEM Cpmsprtium 00/mmg
Plastic Molding Compound
How does exceeding the Tg affect PEM reliability?
Tg’s of mold compounds commonly in use range from 150°C to 165°C.
Exceeding the Tg over time can:
– Breakdown chemical cross-linking of polymers
– Release previously bound up flame retardants and ionics
– Cause corrosion, device instability or lift bonds due to release of ionics
– Reduce temperature cycle capability (due to high CTE)
– Reduce adherence causing delamination
Newer compounds currently under evaluation by Intersil have a Tg of 200°°C.
Potential tradeoffs have to be assessed.
PEM Cpmsprtium 00/mmg
Plastic Molding Compound
How consistent is the molding compound material from lot-to-lot?
Intersil has extensively measured key parameters over several years using
SPC charts.
SPC of Mold Compound Parameters
Parameter
#Lots
%Lots <LCL
% Lots >UCL
Glass Transition Temp (Tg)
464
1.0
1.3
Coeff. Therm. Expansion (CTEa1)
464
1.7
0.8
Coeff. Therm. Expansion (CTEa2)
442
0.0
0.2
Filler Material (% Weight)
442
0.4
0.4
Moisture Uptake (% Weight)
328
1.2
1.2
Note: CTEa1 (40-100°C), CTEa2 (180-220°C)
PEM Cpmsprtium 00/mmg
Plastic Mold Compound
How are impurity levels controlled?
Max.
150
Hydrolyzable chloride, ppm
Moisture Uptake, %
0.5
Water Extractable Species 3
Typical
Levels
Observed
50-130
0.2 - 0.4 Trace Metals
Zinc, ppm
Iron, ppm
Calcium, ppm
Max.
Typical
Levels
Observed
100
300
300
<100
<100-500
1 - 50
---
100 - 1000
1 - 50
1 - 50
1 - 20
1 - 10
1 - 50
1-5
1 - 100
Water Extract Conductivity,
mmho/cm
Water Extract pH
150
3.0 - 6.0
Sodium, ppm
5
Potassium, ppm
5
Chloride, ppm
5
Bromide, ppm
20
Total Halides, ppm
50
30-70 Aluminum, ppm
3.5 - 6.5 Chromium, ppm
0.5 - 3.0 Copper, ppm
Magnesium, ppm
<0.5 - 3.0
Manganese, ppm
1.0 - 5.0 Nickel, ppm
5.0 - 20.0 Silver, ppm
<30 Titanium, ppm
PEM Cpmsprtium 00/mmg
PEM Temperature Range
PEM Cpmsprtium 00/mmg
PEM Temperature Range
Can PEMs meet the Military operating temperature range
(-55°C to +125°C)?
Most PEMs designed for following ambient temperature ranges
(Max Tj = 150C):
Commercial:
0°C
to +70°C
Industrial:
-40°C
to +85°C
Automotive:
-40°C
to +125°C
Some PEMs are rated in the Military temperature range (e.g., Intersil Logic
Families).
Not all PEMs are upgradeable to -55°C to +125°C, but do all Military
applications require this temperature range?
PEM Cpmsprtium 00/mmg
Manufacturing Controls for Reliability
PEM Cpmsprtium 00/mmg
Manufacturing Controls for Reliability
How does the production flow differ between Military and
Commercial products?
Wafer fab:
– Both Military and Commercial product share the same flows, which
employ SPC and reliability critical node controls (see attached chart).
Assembly:
– Assembly flows for hermetic and plastic parts differ due to the nature
of the packaging.
– Both flows employ SPC and critical node controls (see attached chart
for PEMs).
PEM Cpmsprtium 00/mmg
Building In Reliability
(BIR)
Manufacturing for Reliability
Design for Reliability
(MFR)
(DFR)
Reliability
Critical Node
Wafer Level
Reliability
Construction
Analysis
Model
Maintenance
Input
Maverick Lot
Prevention
Integrated Yield
Management
Electronic
SPC Systems
Contamination
Control & Monitor
(SIMS)
(ELYMAT)
ESD
Design For Reliability
Wearout
Characterization
Electron Spin
Resonance
Fast Qual
Devices
Bench Marking
BIR Methods
ATE Fault Coverage
Layout Groundrule
IDDQ, V- Stress,
Inputs
Virtual Test
Plastic Package
Thermo-Mechanical & Moisture
Considerations in Design
PEM Cpmsprtium 00/mmg
Building-In Reliability
ÿ
Process Wearout characterization early in development
cycle leads to the following:
Reliability input to the Design/Layout groundrules
Specification of reliability critical node parameters (Wafer Fab &
Assembly)
Statistical Process Control, Variance Reduction Methods (SPC) and
Problem Solving Methods:
The absence of Wearout Failure Mechanisms in the useful life of
product.
PEM Cpmsprtium 00/mmg
Critical Nodes
ÿ SPC control parameters related to reliability failure
mechanisms.
ÿ Cpk’s have a minimum goal of 1.33 to reliability limits;
screening limits may be used to improve Cpk levels.
ÿ Yearly capability studies are required; capability data can
come from WAT when correlation is shown with in-line SPC.
ÿ Reliability and Manufacturing coordinate the assignment and
implementation of nodes at all Fab sites.
ÿ Data from Critical Nodes is available to customers.
PEM Cpmsprtium 00/mmg
Generic List of Reliability Critical Nodes
Base Reliability Issue
MOS Active Area
Final Gox Thickness
MOS Gate Length
Final Conductor-Si Dielectric Thickness
Final Poly-M1 Dielectric Thickness
Poly-M1 Dielectric Film Stress
Final M1-M2 Dielectric Thickness
M1-M2 Dielectric Film Stress
Final M1, M2, etc. Thickness
Metal Cross-Sectional Area in and around
Aperture & Via
Final M1, M2, etc., Line Width
Metal Film Quality
Passivation Integrity
Passivation Film Stress
Related Failure Mechanism
Critical Characterization
Node
Node
Hot Carrier
Gox Breakdown
Hot Carrier, Gox Breakdown
Dielectric Breakdown
Poly-M1 Dielectric Breakdown
Stress Induced Voids
Stress Induced Voids
M1-M2 Dielectric Breakdown,
Stress Induced Voids
Stress Induced Voids
Electromigration, Stress Induced Voids
Electromigration, Stress Induced Voids
Electromigration, Stress Induced Voids
Electromigration, Stress Induced Voids
Corrosion, Electromigration, Stress
Induced Voids
Stress Induced Voids
Thickness/
Doping
PIT
PEM Cpmsprtium 00/mmg
Plastic Package Critical Node List (PDIP)
Process Flow Step
Critical Node Parameter
Type of Control
Saw
Kerf Width,
DI Resistivity
X Bar - R
Monitor
Die Visual
Die Attach
Die Attach Cure
Visual Quality
Visual Quality
Oven Temperature
Die Shear
Pull Strength
Visual
Temperature
Force
Ball Shear
Visual Q
Acoustic Microscopy
AQL
NP - Chart
X Bar - R
Z - Chart
X Bar - R
NP - Chart
X Bar - R
X Bar - R
X Bar - R
NP - Chart
Monitor
Visual Q
Oven Temperature
Visual Q
Visual Q
Solder Thickness
Visual Q
NP - Chart
X Bar - R
NP - Chart
AQL
X Bar - R
NP - Chart
Wire Bond
Mold
Chemical Deflash
Mold Cure
Trim/Form
Solder Dip
Brand
PEM Cpmsprtium 00/mmg
Reliability Monitors
Comparison of Military Hermetic and Intersil Plastic
Hermetic Military (MIL-STD-883)
Quality Conformance Inspection (QCI)
Description
Sample/
Acc. No.
Frequency
Group B
Resistance to Solvents
Bond Strength
Solderability (8 Hrs. Steam Age)
3/0
22/0
10/0
Each lot
Each lot
Each lot
Group C
HTOL (125°C, 1k Hours)
45/0
1X/12 Months
15/0
15/0
15/0
1X/6 Months
1X/6 Months
1X/6 Months
15/0
1X/6 Months
Group D
1. Physical Dimensions
2. Lead Integrity
3. Thermal Shock (15 cycles)
Temp Cycle (100 cycles)
Moisture Resist (10 cycles)
4. Shock
Vib. Var. Freq.
Acceleration
5. Salt Atm. (24-240 HPS)
6. Internal Wafer Vapor
7. Adhesion of Lead Finish
8. Lid Torque
Note:
15/0
3/0
15/0
5/0
1X/6 Months
1X/6 Months
1X/6 Months
1X/6 Months
Plastic Commercial
Matrix Monitor
Descriptions
Matrix I
HTOL (125°C or 175°C, 48 Hours)
HAST (135°C/85% RH, 48 Hours)
Autoclave (96 Hours)
Thermal Shock (200 Cycles)
Matrix II
HTOL (125°C, 1k Hours)
THB (85/85, 1k Hours)
Autoclave (192 Hours)
Storage Life (150°C, 1k Hours)
Temp Cycle (1k Cycles)
Matrix III
Solderability (8 Hrs. Steam Age)
Brand Adherence
Lead Integrity
Physical Dimensions
Flammability UL-94
SPC Monitored (Eqv. to Hermetic)
Bond Strength
Die Shear
Solderability >4 Hours Steam Age
>8 Hours Steam Age
Sample
Acc. No.
Frequency
45/0
45/0
45/0
45/0
2X/Month
2X/Month
2X/Month
2X/Month
45/0
45/0
45/0
45/0
45/0
1X/Month
1X/Month
1X/2 Months
1X/2 Months
1X/2 Months
22/0
15/0
15/0
11/0
5/0
2X/Months
1X/Month
1X/Month
1X/Month
1X/Quarter
SPC
SPC-Z
Chart
Recording
Recording
1X/Shift
1X/Oven/Cycle
1X/Shift
1X/Week
Mil-Std-883 requires assembly locations to have an additional monitor program to Mil-Std-976 (i.e., Bond Strength/Die Shear, etc.)
which has not been covered by this table.
PEM Cpmsprtium 00/mmg
Conclusions
PEM Cpmsprtium 00/mmg
Conclusions
ÿ PEMs will provide the desired reliability if:
–
–
PEM capability matched to application conditions.
Industry’s best manufacturing practices used.
ÿ System designer must become familiar with:
–
–
–
–
–
PEM use envelope and system environment
Potential failure mechanisms
Supplier’s reliability data
Supplier’s quality systems
Best practices for using PEMs
PEM Cpmsprtium 00/mmg
Conclusions - continued
ÿ PEM use envelope is ever expanding.
ÿ PEM reliability has been successfully demonstrated in many
applications.
ÿ Important to consider qualified suppliers experienced in
Military HSMs and Commercial PEMs.
PEM Cpmsprtium 00/mmg
The End
PEM Cpmsprtium 00/mmg