Reviewing an X-Ray Solder-Void Report

A void percentage is meaningful only when the image, measured region, and calculation method are understood. Two reports can give different values for the same joint if they use different views or analysis settings. Before deciding whether a process is acceptable, establish what the reported number actually represents.

Identify the joint and imaging method

Nordson describes X-ray inspection for hidden BGA and QFN joints and distinguishes transmission, oblique, and three-dimensional techniques. Its application information also discusses background structures that can interfere with void analysis. This is a reason to review the measurement setup rather than treating every X-ray image as equivalent evidence. Nordson: solder-joint X-ray inspection.

Record the board and component identity, side, joint or pad designation, image orientation, and inspection method. Confirm whether the result is based on projected area or another defined measurement. Do not describe a two-dimensional area fraction as a volumetric result unless the method actually supports that interpretation.

Inspect the analysis boundary

• Check that the region of interest follows the intended soldered area.
• Identify overlapping copper, components, vias, or opposite-side joints.
• Review how the software separates voids from background and edges.
• Confirm whether individual void size, total area, or location is being evaluated.
• Keep the original image and the annotated analysis together.

For example, a thermal pad with vias beneath it may produce features that require careful interpretation. A threshold setting that works for a simple isolated pad may misclassify part of this structure. Ask the inspection specialist to explain the segmentation and compare a suitable alternative view where needed.

Apply the correct acceptance requirement

Use the package supplier’s guidance, drawing requirements, and the agreed assembly acceptance plan. A limit applied to one package or joint type should not be copied automatically to another. Consider location and function where the specification requires them; the same total projected area can describe different void distributions.

Connect inspection to process improvement

When results change, compare stencil, paste, board finish, via treatment, component lot, and actual reflow conditions. Use a controlled trial to test a proposed cause. Reducing a reported percentage is not the only objective: the revised process must also preserve the rest of the joint and assembly requirements. A passing void report does not independently prove electrical continuity, thermal performance, or long-term reliability. Retain the inspection recipe and approved criteria with the production record so future reports can be compared on the same basis.

If an inspection recipe changes, retain the earlier settings and compare representative samples before combining results in one trend chart.