Printed Wiring Board Strain Gage Test Guideline Rev 23 - …ect-fsg.com/wp-content/uploads/2015/07/StrainGageguideline.pdf · Printed Wiring Board Strain Gage Test Guideline Rev 23

September 2004 IPC/JEDEC-9704 Printed Wiring Board Strain Gage Test Guideline

[email protected] 1

Printed Wiring Board Strain Gage Test Guideline Rev 23

1 SCOPE This document describes specific guidelines for strain gage testing for Printed Wiring Board (PWB) assemblies. The suggested procedures enables board manufacturers to conduct required strain gage testing independently, and provides a quantitative method for measuring board flexure, and assessing risk levels. The topics covered include: • Test setup and equipment requirements • Strain measurement • Report format

This document assumes a surface mount device; Ball Grid Array (BGA), Small Outline Package (SOP) and Chip Scale (size) Package (CSP) are typical device examples. Discrete Surface Mount Technology (SMT) devices, (e.g. capacitors, resistors, etc.) are outside the scope of the guidelines. 1.1 Purpose Strain gage testing allows objective analysis of the strain and strain rate levels that a SMT package is subjected to during PWB assembly and test. Characterization of worst-case PWB strain is critical due to the susceptibility of component solder joints to strain-induced failures. Although solder grid array packages that use electroless, nickel immersion gold-plated (ENIG) substrates are believed to be more susceptible to strain-induced solder joint failure, excessive strain can result in solder joint damage for all package substrate plating finishes. Strain induced failures include solder ball cracking, trace damage, pad lifting and substrate cracking during board manufacturing and test processes. 1.2 Background Board flexure control using strain gage measurement has proven very beneficial to the electronics industry, and continues to gain acceptance as a method to identify damaging manufacturing processes. However, as interconnect densities have increased and become more fragile, the potential for flexure-induced damage has increased. Many board manufacturers are now required to operate under strain levels specified by their customers or component suppliers. As strain measurement technology has matured, different methodologies have developed. Variations in strain gage methodology inhibit reliable data collection

and prevent data comparison across the industry. This document addresses variations in gage mounting, gage placement, experiment design, data acquisition system variables, and strain metrics. PWB strain measurement includes application of strain gages to the board at specified components, and then subjecting the instrumented board to various test and assembly operations. Test and assembly steps which exceed strain limits are deemed excessive and are identified so that corrective actions can be made. Strain limits may come from the customer, component supplier or internal best known practices. Examples of strain measurement criteria are shown in Appendix A.

Figure 1-1 Lifted pad

By identifying areas sensitive to manufacturing variation, strain gage testing provides insight into the effects of a production ramp. Strain gage measurements become the baseline for future process improvement activities, and quantify the effectiveness of adjustments. Assembly steps that are typically characterized are listed below: 1. SMT assembly process:

• Board depanelization (routing) processes • All manual handling processes • All rework and retouch processes • Connector installation • Component installation

2. Board test processes: • In-Circuit Test (ICT) • Board Functional Test (BFT), or equivalent

functional test

3. Mechanical assembly:

BGA solder ball

copper pad

FR4 board

crack


[email protected] 2

• Heat sink assembly • Board support/stiffener assembly • System board integration, or system assembly • Peripheral Component Interconnect (PCI) or

daughter card installation • Dual In-line Memory Module (DIMM)

installation Assembly processes for different boards and manufacturers vary. Tests such as ICT and BFT are referred to generically in this document; nomenclature can vary at different manufacturing sites. In such cases, apply the same requirements to the equivalent test processes. However, the goal is to characterize all assembly steps involving mechanical loading. Do not constrain testing to the steps listed above, or only to perceived high risk areas. The data from these tests serve as a baseline for future reference. 1.3 Terms and Definitions The definition of all terms used herein shall be in accordance with IPC-T-50 and as defined below. Component Packaged semiconductor device Interconnect Conductive element used for electrical interconnection, e.g., solder ball, lead, etc. Microstrain Dimensionless unit, 106 × (change in length) ÷ (original length) Principal Strain The maximum and minimum normal strains in a plane, always perpendicular to each other and oriented in directions for which the shear strains are zero. Rosette Strain gage containing two or more independent grids for making measurements of strain along each of their axes about a common point. Stacked Rosette Strain Gage Strain gage rosette constructed of grids stacked one above the other about a common point. Strain Dimensionless unit, (change in length) ÷ (original length) Strain-rate Change in strain divided by the time interval during which this change is measured Strain gage Planar metallic foil pattern that is adhered to an underlying surface and exhibits a change in resistance when subjected to a strain. Strain gage element Sensing area of strain gage defined by the serpentine metallic grid pattern.

1.4 Future Studies Continued research on this topic will focus on the refinement of prescribed strain limits, as well as research on dynamic impulse events. e.g. drop and shock for bare PWBs, assembled systems and shipping packages. This is important as there is increasing evidence of strain induced BGA fractures that are induced outside of the manufacturing and test process. Printed Wiring Board Assemblies (PWBA) are just as susceptible during shipment. The presented guidelines will continue to be refined based on this research. 2. APPLICABLE DOCUMENTS The following documents are applicable and constitute a part of this specification to the extent specified herein. Subsequent issues of, or amendments to, these documents will become a part of this specification. Documents are grouped under categories as Institute of Interconnecting and Packaging Electronic Circuits (IPC), Joint Electron Device Engineering Council (JEDEC), American Society for Testing and Materials (ASTM) and others depending on the source. 2.1 IPC Documents • IPC-TM-650 Test Methods Manual • IPC-D279 Design Guidelines for Reliable Surface

Mount Technology Printed Board Assemblies • IPC 7095 Design and Assembly Process

Implementation for BGAs • IPC 9701 Performance Test Methods and

Qualification Requirements for Surface Mount Solder Attachments

• IPC-JEDEC 9702 Monotonic Bend Characterization of Board-Level Interconnects

• IPC 9703 Mechanical Shock Test Methods and Qualification Requirements for Surface Mount Solder Attachments

2.2 ASTM Documents • ASTM E1561-93 (Reaffirmed 2003) Standard

Practices for Analysis of Strain Gage Rosette Data 2.3 Other Reference • Code of Practice, for installation of electrical

resistance strain gauges, British Society of Strain Measurement (www.bssm.org)


[email protected] 3

3 GENERAL REQUIREMENTS/GUIDELINES An overview of the strain gage measurement process for both board and system assembly is depicted in Figures 3-1 and 3-2.

Figure 3-1 Board assembly strain measurement

Figure 3-2 System assembly strain measurement


str

3.1 Boards Due to the limited mechanical strain applied prior to SMT reflow, and more importantly because solder joints are formed only after reflow, strain characterization is required only for assembly and test operations following SMT reflow. Typically, a minimum of two test boards are instrumented. They are not required to be electrically functional but must mechanically represent the latest design. At a minimum, evaluate the following two board types: • Board with SMT components only (after SMT

reflow) • Boards with both SMT and thru-hole components

(after wave solder)

These are the minimum requirements. Characterization of the system assembly process might require additional test boards. The first instrumented board reflects a PWBA that has been through SMT reflow, just prior to wave solder. At this stage, the board contains only SMT components. The objective at this stage is to characterize the strain/strain rate during manual handling, insertion of connectors and other thru-hole components, and any electrical testing conducted prior to wave solder. This board should not be used for the characterization of assembly steps after wave solder. The second instrumented board should be similar to PWBAs that have completed wave solder. This board contains all SMT and thru-hole components and is used to characterize all assembly steps after final reflow including (where applicable):
Figure 3-3 Board with SMT components only(after SMT reflow)
[email protected] 4

Figure 3-5 Instrumented board (pre thru-hole component insertion)

• Depanelization /routing • Board support/stiffener assembly • Final system assembly • PCI card insertion • DIMM module insertion • Daughter card insertion • Heat sink attachment • Test operations (ICT, BFT) • BGA and through-hole component rework Although ICT and BFT are typical high strain/strain rate operations, damage is possible in any other step. All assembly steps should be characterized. Attention should also be paid to processes where mechanical fixtures are used. e.g. support fixtures, press fit fixtures, thumbscrew fixtures, etc.

Figure 3-4 Board with both SMT and thru-hole components (after wave solder)


[email protected] 5

It is strongly recommended that any manual handling between assembly steps, with or without fixture, be characterized. If the manual handling steps are similar, combining the handling test into one test run representative of worst-case handling is acceptable. Details of this manual handling simulation must be documented in the test report. Simulations should also be conducted to quantify the associated variability. There could be unique manufacturing processes that require alternative configurations. For example, thru-hole components typically require wave solder. Wave solder conventionally follows convection reflow (one or two passes depending on board layout). However, a PWB could have inductor coils manually inserted before SMT reflow and not require wave solder. In cases where assembly characterization prior to wave solder is required, the test board must be mechanically representative of boards prior to SMT reflow. In such instances, alternative set-ups are acceptable as long as all mechanical loading characterization requirements are met. Components of large physical size and/or mass must be also present on the board. It is recommended that test boards be inspected for excessive warpage prior to instrumentation. 3.2 Components and Devices It is recommended that any BGA device with a package body size equal to or larger than 27mm x 27mm be measured. For example, Tape Ball Grid Array (TBGA), Flip Chip Ball Grid Array (FCBGA), Enhanced Ball Grid Array (eBGA) and low standoff BGA components that meet this requirement should be characterized. This includes electrolytic Ni/Au-plated substrates (typical for wirebond packages), even though these are less susceptible to solder joint failure than ENIG-plated package substrates (typical for flipchip BGA packages).

If all BGA devices on the PWB have a package body size smaller than 27mm x 27 mm (e.g. handheld devices, cell phones, etc.), then the largest 3 BGA components, at a minimum, should be measured. Alternative criteria to determine strain gage placement includes observed failure locations, historical failure rates, finite element analysis, assembly/test fixture configuration, board design, and BGA package design. Board and package design considerations include geometry, materials and configuration. Failure locations may be identified using dye (Dykem or equivalent) penetration and component removal ("dye-and-pry") techniques. For boards with a larger number of BGA components (i.e. 6 or above) it is acceptable to rely first on a FEA model, or other analytical and computation methods, to predict the optimized placement of strain gages. However, if initial testing identifies areas of high strain, this will be followed up with more focus testing in the subject area. Figure 3-6 ICT fixture lead wire routing 3.3 Strain Gage Details of the recommended strain gage are as follows:

Figure 3-7 Stacked Rosette strain gage • 3 element stacked rectangular (0/45/90) rosette

strain gage • 1.5 x 1.5 mm sq, nom., gage sensor size • 120 Ω or 350 Ω strain gages • Lead wire attach pads located at or lead wires

attached on one side of strain gage The gage length should be as small as possible to minimize effects of non-uniform PWB strain gradients. However, strain gages should be large enough so that small features such as traces and vias, do not affect the strain reading. For strain measurements at varying temperatures, the Coefficient of Thermal Expansion (CTE) of the strain gage should be matched to the PWB substrate.


[email protected] 6

Strain gages with or without pre-attached lead wires can be used. Selection should be based on preference and specific applications. Strain gages with pre-attached lead wires have the advantage of not requiring lead wire soldering, but it can be more difficult to maintain a high quality epoxy bond line during strain gage attachment. Lead wires can be re-soldered but the bond line cannot be re-done. Conversely, manually soldered lead wires could result in electrical shorts. Lead wire soldering is best performed under a 20-50× optical inspection microscope. Local bending near components will create variations in gage readings. For this reasons gage placement must be precise. Wherever feasible the gages should be placed as described below. In cases where the preferred placement is not feasible, alternative placements selection should be made in concert with the customer and may require modification of the board assembly. The preferred gage placement for gages is to have strain gages on all fours corners of the selected component, with the centers of the rosette (not the rosette backing matrix) located at the intersection of lines offset from the package edge 0.14” +/- 0.01” (3.56mm +/- 0.25mm), as shown in Figure 3-9, below.

Grid strains e1 and e3 should be oriented parallel to the edges of the package e2 should be oriented diagonally away from package, with respect to the edges of the package. The consistent and precise placement of gages is critical to correlation of data between test location and samples.

The distance between the gage and the BGA components might vary at each corner due to varying constraints. In such instances, gages should be placed as close to each BGA package as possible. This information should be indicated in the test report, including photographic documentation. There may be situations where strain gages are not necessary on all corners of a device, such as when two or more devices are immediately adjacent or banked. In such instances, employ analytical techniques or computational models to identify the locations of the highest strain, thereby reducing the number of required strain gages. All supporting assumptions and analysis must be clearly documented in the test report. There may also be situations where strain gage placement is limited mechanically. For example ICT fixtures have ICT probes, pushdown pins and BGA pushdown blocks which prevent strain gage placement in desired locations. In such instances, removal of part of the component may be considered as an alternative. When removing the corner of any component, the gage should be placed so that its centroid is placed on top of the corner land pad on the PCB. The removal of the component should be limited to what is necessary to facilitate the placement of the strain gage. In the event that two gages at the corners of adjacent components would overlap, the test should be conducted using multiple boards. It is suggested that

Offset from the package edge

0.14”

ee33

ee22ee11

Figure 3-9 Recommended gage placement

Figure 3-8 Strain gage dimensions (inches)


[email protected] 7

all corners of the component be gaged on the same board. 3.4 Gage Attachment Board preparation is a critical part of the instrumentation process. Proper board preparation will help ensure that proper bonding of strain gages, this will in turn improve the accuracy of readings. Recommendation for board preparation and attach the strain gages follows: Prior to strain gage placement, prepare the surface to ensure proper adhesion of the bonding epoxy. 1. Desolder small components and discrete

components that interfere with gage placement. 2. Roughen the surface using sandpaper, e.g. 400 grit

silicon-carbine paper, or similar.

3. Clean the surface with a solvent, such as Isopropyl alcohol. Solvents used MUST be chemically clean.

4. Once the surface has been prepared, attach the strain gages using epoxy.

If attachment of strain gages to the top side of the board is not technical feasible, attachment to the bottom side of the PWB is acceptable. Considerations for bottom side strain gage placement follow: 1. Placement should be directly underneath the

subject module corner. 2. Where possible, FEA models should be use to

determine the relationship of top side board strain to bottom side in the analysis of the results.

Strain gage attachment is performed in accordance with instructions provided by the strain gage and epoxy suppliers. Note that strain gages require the use of specially formulated epoxy. For details, check with your strain gage supplier.

Figure 3-10 Component removal to facilitate gage placement

3.5 Lead Wires Details of the suitable lead wires are as follows: • 30 American Wire Gage (AWG) lead wire

preferred; Poly(vinyl chloride) PVC or Kynar insulation, OR;

• Single solid copper wire with polyurethane enamel coating

• 3-wire configuration (allows lead wire resistance compensation) preferred over conventional 2-wire configuration

• 1.5 to 2.5 meters lead wire length recommended, but wire length should be no longer than needed.

Figure 3-11 Centroid of gage placement above corner land pad. A two wire quarter bridge doubles the desensitization

of the strain gage, can introduce a significant amount of temperature sensitivity just due to leads, and creates a potential balancing issue for the instrumentation. For the most stable static measurements, a three wire system should be used. Actual selection of lead wires may vary depending on the specific application. For instance 30 AWG wire may not be useable for the top side vacuum seal in an ICT fixture. As lead wire routing is typically most constrained in ICT fixtures, lead wires must be routed in such a way as to avoid interference with supports and push-down posts when the fixture is engaged.


[email protected] 8

If the same test board is used for both ICT and BFT, lead wire routing must accommodate the footprint of the mechanical support and pins for both fixtures. Single solid copper wires can also help facilitate routing in ICT fixtures, and also help minimize vacuum leakage. Alternatively, ICT fixtures can also be designed to better accommodate the routing of lead wires. Some considerations are presented in Appendix C.

Figure 3-13 Lead wire routing example

Reinforce the lead wire attachment at each strain gage with epoxy or tape. Use adhesive-backed polyimide film or fiberglass cloth. In some cases, strain gage lead wire routing on the top side of the PWB might be impracticable due to interference from heat sink supports. Attaching the strain gages to the bottom of the PWB might be the only viable option. Locate strain gages immediately adjacent to the outer BGA solder ball rows on the PWB bottom, similar to top-side gage placement.

Bottom strain gage attachment is an exception and is not recommended. Only use it when top side attachment becomes excessively restrictive. 3.6 Measurement Equipment Strain measurements are sensitive to scan frequency, data signal bit width, and strain signal limits (gain). All sampling must be simultaneous, as sequential sampling may result in miscalculated strain values. The following guidelines are recommended: • For high impulse events, a minimum scan

frequency of 500 Hz is recommended. e.g. ICT, drop.

• For general low impulse assembly processes, a minimum scan frequency of 100 Hz is recommended. e.g. in-process handling, rework, etc.

Figure 3-12 Lead wire routing example

A minimum sampling resolution of 12 to 16 bits is recommended. Adjust signal amplifier gain for optimum use of dynamic range (that is, maximize the gain, but set it low enough to prevent clipping of peak strain values) If the data appears truncated, the measurement frequency should be increased to verify there is not a high frequency dynamic event occurring. Additionally, the number of available monitoring channels limits the number of measurements in any one pass. While multiple passes are allowable if there are insufficient channels, all three gages in any stacked rosette must be monitored at the same loading. Since the PCB material has low thermal conductivity, gages are more likely to heat up due to the electrical current passing through them. While making use of a three lead wire setup and quarter bridge will reduce this effect, the excitation voltage should be balanced with the signal/noise ratio. If the strain value drifts significantly while the board is at rest, the voltage should be reduced until this effect either disappears, or the signal/noise ratio becomes unacceptable. In general, an excitation level of 2V should provide satisfactory performance. 3.7 Measurement Calibration As many of the procedures above can lead to errors in measuring the board strain, a simple calibration jig has been developed. This jig can be used to find and eliminate many errors due to gage placement, gage attachment, data acquisition system setup, and lead wires. This jig is shown in Figure 3-14. In this jig, a PWB coupon board is deflected by steps by the insertion of shims.


[email protected] 9

The strain for each shim is recorded and can be compared against expected variation. The fixture is only meant to check for basic errors in gage attachment and measurement. It can not be used to capture dynamic (sampling rate) errors, nor does this check for errors associated with the actual manufacturing equipment, i.e. thermal effects, or wire interference with manufacturing equipment. 3.8 Manual Simulation Apart from mechanically actuated test operations, most other assembly steps are manually simulated. This is an integral part of strain gage testing. The results from these simulations can be grouped into two general categories: • Observed handling • Excursionary handling Observed handling represents handling that PWBs would typically experience during assembly and test. Strains imposed would be representative of nominal loads exerted during assembly and test. Excursionary handling on the other hand represents extreme or worst case events. While these events are unlikely, they represent a real and potential risk and should be understood, e.g. accidental drops, improper handling, etc. Such testing is important as it will determine if the subject is susceptible to such excitations. The recommended procedure for manual handling simulation is described in the following paragraphs: Good judgment must be exercised to ensure that simulations are representative of the worst-case assembly strain profile. Simulation of manual handling and failure analysis steps should be based on observation of actual work practices, not on a volunteer assembler.

Repeat each step three times to capture the natural variation of the subject assembly step. Do the same for mechanically actuated test steps. For instance, to simulate thru-hole connector insertion prior to wave solder, press the actual connector to simulate the actual insertion process. At a minimum, the following simulations are recommended to simulate in-process handling: • Drop on surface • Load while suspended • Bounce in air 1. By holding the PWB on one edge, position it 10 to

15cm parallel to a rigid surface, with the top side facing up. a) Allow the opposite edge to fall to the surface

to simulate the accidental drop of the PWB during manufacture.

Figure 3-14 Gage correlation tool b) Repeat for all four edges – once with the top side of the board facing up and another with the top of the board facing down.

2. With the board suspended on actual or simulated

edge guides, apply a 5 to 10kg load for approximately two seconds at all component locations instrumented with strain gages, and release.

Figure 3-15 Simulated edge guides

This simulates both intentional and unintentional loading of the PWB during manufacture. Perform the prescribed action on both the top and bottom of the board. If the manufacturing setup is different, a more representative setup should be employed. 3. Hold the PWB by one corner.


[email protected] 10

a) Move the board approximately 5cm upwards and downwards in the air, in a quick “bouncing” motion.

b) Repeat the prescribed action for each corner – once with the top side of the board facing up and another with the top of the board facing down.

The steps above should only serve as guidelines. In all applications, exercise judgment to identify additional simulations. Unless strict procedures absolutely prevent occurrence, all possible worst case handling processes should be simulated. Particular attention should be paid to the characterization of manual handling with unintended hand-hold locations. e.g. heat sinks, board stiffeners, supports, etc. 3.9 Shipping Package Test The assembly line is not the only area where BGA fractures can be induced. PWBs are also susceptible during shipment. Strain gage testing can be used to complement and to provide further insight into the effectiveness of the package, and the strain that components are subject to. The purpose of the drop test is to subject the shipping package to specified test conditions, in order to determine if the box would allows excessive board flexure, which could result in cracked BGA issues. For this purpose, one or two boards instrumented with strain gages are placed in the front slots of the shipping box to measure board flexure during the specified test test.

Instrumented boards should have SMT and through-hole components. Similar guidelines would apply for such testing, however due to the impulse nature of the forces applied, a higher sampling frequency would be required to ensure that strains are accurately measured. i.e. drop testing. A minimum sampling frequency of 500 Hz is recommended.

Figure 3-17: Placement of instrumented boards for shipping package test

4 DATA ANALYSIS AND REPORTING Generate a test report once data collection is complete. The recommended report format is represented in Section 4.3. 4.1 Analysis Requirements The details of the analysis will vary with the particular strain limit criteria being employed. At a minimum the peak values of the principal strain (maximum and minimum) should be given for each step monitored. Other strain limit criteria may require calculation of the strain rate. Strain rate calculations may use a least squares fit or similar technique over the range of interest to avoid sampling errors. For operations that contain multiple steps (i.e. system assembly) it is recommended that the time history of the strain limit criteria be plotted as in Figure 4-1. This will assist in making any needed changes to the operation. Analysis data must highlight high risk operations that exceed the strain limit criteria. Any high-risk areas, as defined in this section, should be further analyzed.

Figure 3-16 Shipping package drop test


strai

4.2 Tfollo •

•

•

• • In apart procprogare w All be tdiffe 4.3 Sshal 1. Onepass 2. Onedesc 3. Detathe t

• Test board (Include information on components characterized, i.e. package type, solder ball pitch, etc., and board thickness.)

DIMM Insertion

HD Connectors

Peripheral connectors (i.e.. monitor, mouse,

keyboard, etc…)

PCI Connectors

Video 2-pin RM

• Strain gage specification • Strain gage placement information, i.e. the

distance of the X-Y offset from subject component.

• Strain measurement equipment • Details of each assembly process (such as

revision and serial number of fixtures, if applicable)

4. Results Detailed summary of test results using words, tables, and graphs. If the experiments were used to determine

Figure 4-1 Time history of the strain limitcriteria


est Frequency Conduct strain gage testing in the wing events:

Before any test or assembly fixture is brought online Whenever there are modifications to a test fixture that might alter the strain profile Whenever there are modifications to an assembly process that might alter the strain profile In-process stiffener design validation Enclosure design validation prior to hard-tooling

ddition, strain gage tests should be conducted as of the routine Preventative Maintenance (PM)

ess. It is strongly recommended that a PM ram be developed to ensure that the strain profiles ithin specified limits at all times.

test fixtures, including spares and backups, should ested. Fixtures of identical design can possess rent strain and strain rate profiles.

train Gage Test Report Template Test Reports l be in the following format.

Abstract paragraph executive summary of results. Use a /fail table to summarize the results.

Introduction paragraph explanation of test purpose, and general ription of test.

Test Apparatus and Setup iled description, using words and photographs, of est equipment.

• Date of test

the optimal setup or design, clearly define the experimental process and analysis outline. Compare results to historical data where available. 5. Conclusions Detailed summary describing the most important experimental results and list of recommendations for either changes in board assembly procedures, design changes for the system or test fixture, or further testing. Clearly explain justification for the recommendations. 6. Future Studies One paragraph description of recommended future testing.


[email protected]

APPENDIX A: STRAIN LIMITS Three strain limit approaches are proposed. The first is a component-specific approach, in which strain limits and specific gage locations are provided by the component manufacturer. The second approach is a customer specified strain limit against which all components are compared. The third is a similar to the first two, except the strain limit varies with strain rate. The approaches below are based on the assumption that the strain limits will be specified to the board manufacturer by either the customer or component supplier. In the absence of these specifications the board manufacturer may need to develop an internal strain limit criteria based on measuring the strain induced by known good and known bad (damaging) processes. Component Supplier The component-specific approach is based upon the premise that the strain limit is dependent upon variables such as solder ball pitch, package size, package stiffness, solder material, and board thickness. The component manufacturer is responsible for defining their testing technique, gage location and orientation, and strain metric, and providing all of these to the customer. This limit should be based upon the strain level at which damage begins, rather than electrical open. Customer Specified In this approach, a strain limit is determined that applies to all components as specified by the customer. In many cases a single strain limit applies to may be applied to all components on the board. This limit is based upon either testing or empirical data, which defines a critical strain value. A safety factor is usually applied to this value in order to set the limit. Rate Limited The allowable board strain has been observed to be a function of the strain rate or the rate at which the board is bent. The strain rate used is the rate at or near the highest strain level, as shown in figure A-1. In this case, Region C should be used to calculate the strain rate, as it correlates with the highest strain level. A least squares fit should be used to fit a line to the region, the slope of which is the strain rate to be used to determine the strain limit.

1

0

100

200

300

400

500

600

700

800

900

000

4.5 4.6 4.7 4.8 4.9 5 5.1 5.2 5.3 5.4 5.5

Time (seconds)

Mic

rost

rain

Region A

Region B

Region C

Figure A-1 Calculation of strain rate for a typicalICT test

12

Using the method described above the data pairs of strain and strain rate may be plotted against the provided as shown in Figure A-2 on the following page. Further reference for this method can be found in Appendix B.



Figure A-2 Example of allowable strain as a function of strain rate and board thickness



APPENDIX B: REFERENCE FOR RATE LIMIED GUIDANCE Assumptions The guidance in Figure A-2 for maximum allowable PWB strain is based upon empirical studies, finite element modeling and analysis of production failures; however, there are numerous limitations, assumptions and simplifications associated with the graph. The empirical data is largely based upon 35-45 mm BGA devices that used ENIG substrate plating, eutectic SnPb solder, and were assembled on circuit boards with a Hot Air Solder Level (HASL) surface finish, and a 2-3 mm board thickness. PWB strain measurements typically were taken immediately adjacent to the corner-most package solder joints, using 1.5 mm x 1.5 mm nominal stacked rosette strain gages, and recorded at a scan frequency of 100 Hz. Finite Element Analysis (FEA) assumed linear-elastic material properties, orthotropic PWB (FR4) properties, eutectic SnPb solder composition, and symmetric loading. Material properties used for the abovementioned FEA are as follows:

Solder

EX 2.8e6 psi ALPX 24.5e-6 ppm/deg C NUXY 0.36

PWB Resin

Ex 2.8e6 psi Ey 0.5e6 psi

Gxy 0.2e6 psi Gyz 0.2e6 psi Gxz 1.0e6 psi

Alpx 14.5e-6 ppm/deg C Alpy 67.2e-6 ppm/deg C

Nuxz 0.16 Nuxy 0.25 Nuyz 0.25

Copper

EX 18.7e6 psi ALPX 17.5e-6 ppm/deg C NUXY 0.34

The solder joint fracture dependencies of spherical bending versus orthogonal loading are ignored. Additional variables such as component surface finish, PWB surface finish, pad geometry, external heatsink, SMT profile, flux chemistry, etc. that also could affect solder joint fracture resistance are not accounted for in the acceptance criteria.

Analysis Requirements Generate a strain against strain rate plot for the detailed analysis. Figure B-1 is an example of such a plot. Definition for strain rate is documented in Section 1.3. Compare the plot with the recommended acceptance criteria. Refer to Appendix A (Figure A-2). Analysis data must highlight the following conditions: • Strain/strain rate combinations that exceed the

recommended guidelines • Gage readings at one package corner that are more

than double the other corners, when gage readings are above 100 µstrain. (That is, 100% higher than other values)

• Strain/strain rate combinations that are within 20% of maximum allowable guidelines

Any high-risk areas, as defined in this section, should be further analyzed.


[email protected]

Figure B-1 Data Reporting Example for 2.35mm Thick PWB

15



APPENDIX C: ICT DESIGN CONSIDERATIONS To insure a properly designed fixture, a Computer-Aided Design (CAD) file of the Unit Under Test (UUT) nomenclature is necessary. The CAD file is created from a PCB design package (it is recommended to use the PCB design output since it typically will have more detailed information than photo plot files, such as Gerber files). The CAD file is used by the test fixture manufacture to accurately place UUT supports, pockets for UUT components, and placement of strain gauges. Output from PCB design packages can vary. The output results depend on the input of board and component information. Because of this, and because of the number of design packages used throughout the industry, it would be difficult to standardize the parameters of the output file. As a general guide, the more detailed the input information is, the better the output file will be. When designing for strain gauges on a UUT, the test fixture manufacture must put emphasis on the function of the fixture and supporting the UUT. During design of UUT supports, it would not be rational to reduce the number of necessary supports to allow for placement of the strain gauges. The first priority must be to support the UUT, then to determine where the strain gauges can be mounted. Once the UUT supports are designed, the strain gauges can be placed during the remaining design of the test fixture. At this point, the test fixture manufacture may be able to provide paths or slots to allow clearance for the strain gauge wires. However, in many cases, the strain gauge placement is determined after the fixture build. Physical artwork of the UUT supports may be laid across the UUT, showing where the strain gauges can be mounted and insuring clearance for the strain gauge wires.



APPENDIX D: GLOSSARY ASTM American Society for Testing and Materials AWG American Wire Gage BGA Ball Grid Array BFT Board Functional Test CSP Chip Scale Package CTE Coefficient of Thermal Expansion CAD Computer-Aided Design DIMM Dual In-line Memory Module ENIG Electroless, Nickel Immersion Gold-plated eBGA Enhanced Ball Grid Array FEA Finite Element Analysis FCBGA Flip Chip Ball Grid Array HASL Hot Air Solder Level ICT In-Circuit Test IPC Interconnecting and Packaging Electronic Circuits JEDEC Joint Electron Device Engineering Council PCI Peripheral Component Interconnect PVC Poly-vinyl chloride PM Preventative Maintenance PWB Printed Wiring Board PWBA Printed Wiring Board Assemblies SOP Small Outline Package SMT Surface Mount TBGA Tape Ball Grid Array UUT Unit Under Test