3D Solder Paste Inspection

3D SPI Measurements and Inspection Limits

3D SPI measures the geometry and position of every exposed solder paste deposit and compares the results with validated inspection limits. Therefore, valid SPI inspection measurement limits must come from stencil-aperture theory, pad geometry, stable production data, measurement capability related to environmental factors, and customer specification requirements.

Paste Volume Against Stencil-Aperture Theory

During the development of an aperture family, the gap between the theoretical aperture volume and the actual paste volume deposited on the pad is an important diagnostic measurement. It also indicates expected process loss when it remains stable. When the difference between these two values grows, the gap indicates that a potential release problem is developing and that action should be taken before the problem appears as a limit violation. Paste transfer efficiency—the volume deposited on the pad divided by the theoretical aperture volume—provides a useful way to compare the deposited paste with the intended deposit without treating all aperture families as identical.

The factors that affect release include aperture area ratio, stencil thickness, paste rheology, surface finish, aperture treatment, and package geometry. Therefore, SUGA establishes separate baselines for relevant pad or aperture families and uses FAI verification and stable production data to confirm that the patterns and tolerances remain valid. Production limits are tightened or widened only when stable data and measurement capability support the change.

Paste Geometry Beyond Volume

The area, height, volume, offset, and shape of a paste deposit address different questions. For example, two paired pads may have acceptable individual volumes but still show an imbalance associated with tombstoning risk. The same directional offset can have different implications depending on the accompanying measurements: with unchanged height, it may indicate registration error or stencil stretch, while a local height shift is more likely to indicate board movement. Therefore, the disposition rule should use a combination of measurements rather than any single measurement in isolation.

Setting 3D SPI Inspection Limits

Limit setting for 3D SPI inspection starts with the intended deposit and then uses FAI results, measurement capability, stable process distributions, customer specifications, and package-specific risk factors to define acceptable production limits. The program must retain the stencil revision and paste condition, along with the pad or aperture family used to determine directional offsets. It must also retain any paired-pad logic used to reach a decision.

J-STD-005B provides guidance for the qualification, characterization, testing, and classification of solder paste. IPC-7525C provides guidance for stencil design, IPC-7527 provides guidance for solder paste printing, J-STD-001J provides requirements for the soldering process, and IPC-A-610J provides acceptance criteria for assembled products. However, none of these documents provides one universal SPI volume percentage for every pad and package.

Customer drawings and process specifications are converted into board-level limits only after the measurement system is capable of distinguishing the required change. Limits that reject a stable distribution create false calls, while limits that accept a genuine shift allow escapes. The defined window should therefore be supported by repeatability data and a homogeneous population of pads or apertures rather than one percentage applied to the entire board.

How Does SPI Fit Into Assembly Inspection?

SPI is a tool used to verify the printed paste deposit before component placement. Subsequent inspection stages, including AOI and X-ray, assess the physical placement of components, the quality of visible solder joints after reflow, and the internal structure of solder joints.

Defect Visibility by Inspection Stage

Once a component is placed on top of a pad, an optical inspection system cannot recover the condition of the original printed paste surface or its geometry before component placement.

SPI cannot verify component polarity, the quality of the solder fillet after reflow, or the internal structure of the solder joint between a BGA or QFN device and the PCB assembly. Pre-reflow AOI checks component presence, intended location, and orientation. Post-reflow AOI checks the condition of components and visible solder joints. X-ray inspection is used in addition to AOI when the required evidence is inside or beneath a package.

What Inspection Follows SPI?

After SPI, pre-reflow AOI can check component presence, intended location, and polarity after placement. Post-reflow AOI checks the condition of components and the quality of visible solder joints. BGA, QFN, and other internal solder joints are checked using X-ray inspection in addition to AOI when required.

Although manual assessment is not a replacement for repeatable pad-level volumetric measurement of solder paste at production rates, it is useful for FAI, unusual contamination, ambiguous machine calls, and rework assessment.

Configuring 3D SPI for Production

The setup for 3D SPI production starts with selecting the SPI configuration based on production specifications such as minimum pad and gap sizes, board envelope, warpage, cycle time, measurement capability, data interface, and line conditions. The qualified production program uses a single configuration instead of blending limits from multiple configurations across different systems.

2D or 3D SPI?

When the disposition of the PCB relies on measured height, integrated volume, three-dimensional shape, paired-pad balance, or local-surface compensation, the SPI solution needs to be 3D. On the other hand, if evidence of area, presence, and position is sufficient, 2D SPI may suffice. The decision to use 3D or 2D SPI should be based on an assessment of defect risk, pad geometry, board movement, and the production control plan. The package name alone should not determine the selection.

How Is SPI Measurement Capability Proven?

Each SPI measurement capability is established as follows: resolution is defined as the smallest increment that can be reported or distinguished. Resolution should not be confused with accuracy. Accuracy is determined by comparison against a controlled, traceable reference, whereas repeatability is established by measuring the same target multiple times using the same equipment under controlled conditions. GR&R evaluates measurement variation against the manufacturer’s decision limits, including the defined sources of reproducibility.

Calibration targets are used to verify the accuracy of measurement systems. Repeated testing of the same target verifies SPI equipment repeatability, while GR&R evaluation demonstrates whether the SPI measurements are adequate for production purposes. When performing a production study, representative boards and pad families should be used. The study should record the optics, recipe, paste condition, and environment, including temperature, humidity, and elapsed time. It should also include pad geometry, tolerance, sample plan, and whether the GR&R value is reported as a percentage of study variation or tolerance. A GR&R value below 10% has little meaning unless the denominator and actual decision limits are provided.

Balancing SPI Resolution With Cycle Time

High-resolution imaging improves the measurement of deposit edges and small deposits but increases the number of fields, projections, and calculations required to inspect a single board. Resolution, field of view, projection count, and cycle time need to be qualified as one production configuration instead of being selected from unrelated maximum specifications.

Configurations should be compared in seconds per production board, including board loading, fiducial search, selected resolution, projection count, pad count, board transfer, review calls, and final disposition rules. The actual cycle time for a production board determines whether the SPI system supports the line’s takt time.

Handling Warped PCBs

The selected SPI configuration should allow safe transport of the actual PCB without transport or support movement being interpreted as a false height signal. Several variables, including board envelope, thickness, weight, rail margin, underside clearance, carrier design, local flexibility, and the verified compensation range of the system, should be examined together on the production board.

The production trial should include rail engagement, top and bottom clearance, transport height, support tooling, local flex zones, and repeated passes through the conveyor. Thin boards and unsupported transition zones can generate more datum movement than a large, rigid assembly.

Factory Software Compatibility

Verify the actual software license, machine version, report schema, board identity field, MES mapping, and printer-feedback message required by the selected line. The mention of an interface name does not confirm that the required alarm, pad result, recipe revision, or disposition field is available. The message and recovery process must be tested with the production configuration.

Controlling False SPI Calls

False-call performance is influenced by optical parameters, surface compensation, measurement stability, region programming, pad-family limits, and review rules. Controls for false rejects and false accepts should be established using calibration, repeatability studies, representative boards, FAI distributions, multi-angle imaging where applicable, threshold analysis, and confirmation of recurring exceptions before limits are changed.

The review record should include the original machine call, the associated image or height-map evidence, the reviewer’s decision, the reason code, the program revision number, and the final board disposition.

How Early SPI Detection Reduces Manufacturing Cost

A failed board at the print stage can often be cleaned, printed again, and reinspected. A failed board after placement or reflow may require component removal, pad cleaning, joint rework, electrical diagnosis, containment, or a line interruption. The greatest savings occur when paste-related defects would otherwise be identified after valuable components or completed assemblies have entered the manufacturing process.