Start SMT troubleshooting with the observed condition and its record. A defect name describes what was seen; it does not prove why it happened. Preserve board identity, component reference, build revision, inspection images, station results and any intervention before changing the setup.
The checks below are investigation prompts. Use the approved product requirements and inspection method to decide acceptance. Do not adopt a generic temperature, void percentage or rework instruction as a substitute for that decision.
Build an evidence trail before adjusting the process
| Observed condition | Evidence to collect | Checks to sequence |
|---|---|---|
| One end of a component lifted | Reference, orientation, inspection image, pad/land pattern, paste observations and relevant process record. | Confirm the condition, compare the intended geometry and print evidence, then review placement and thermal history with the process owner. |
| Bridge or solder ball | Exact location, image, adjacent features, print record and whether the condition repeats by reference or panel position. | Check the released pattern and stencil/print inputs before comparing placement and reflow conditions. Retain the original sample. |
| Suspected insufficient solder | Inspection method, accessible views, fitted part identity, pad geometry and available paste data. | Confirm the reported criterion and relevant views; compare print, placement and process evidence before deciding repair or process change. |
| BGA or thermal-pad void indication | X-ray setup, image, component/reference, specified evaluation method and related assembly record. | Establish the applicable criterion and measurement method. Relate the indication to print and process evidence; do not infer an electrical failure from the image alone. |
| Offset, polarity or wrong component | Assembly drawing, BOM and placement revisions, component marking, setup verification and first-article record. | Reconcile the intended configuration first, then check setup and feeder/placement records. Hold conflicting instructions for engineering disposition. |
Distinguish a board issue from a measurement or fixture issue
Record the inspection program or test procedure revision and how the result was reproduced. A changed threshold, camera setup, fixture contact or firmware configuration can alter a result. Compare the same conditions and retain both the original and repeat records. The test-fixture handoff guide lists the information needed to make that comparison reviewable.
For hidden-joint indications, use the BGA and thermal-pad void review guide to define the imaging and evaluation questions. X-ray findings and functional-test results answer different questions; one should not silently replace the other.
Change a controlled input and record what changed
Describe the proposed change, the hypothesis it tests and the comparison that will be made. Keep the affected revision, sample scope and process conditions with the results. If several inputs change together, record that limitation rather than attributing the outcome to a single adjustment.
Possible investigations may concern land pattern, stencil, print setup, placement, thermal process, material condition or inspection settings. The appropriate owner should select the method for the actual assembly. The panelization guide is relevant when location, support or depaneling conditions need review.
Keep a compact investigation log
| Record | Example of the information to enter |
|---|---|
| Observation | Reference designator, unit/lot identity, observed condition and evidence file. |
| Hypothesis | The proposed explanation and the evidence that would support or contradict it. |
| Intervention | Changed input, authorization, revision and affected sample scope. |
| Result | Actual measured or inspected outcome, method and limitations; leave blank until the check is performed. |
| Disposition | Approved action for held material and the verification required for subsequent builds. |
Verify the action against the original problem
Repeat the relevant check under the agreed conditions and retain the result. Review whether the action creates another assembly or inspection concern. Identify what happens to held units and whether the change applies to a single lot or future production. A repair that makes one unit pass is not sufficient evidence that a recurring process problem has been resolved.
When similar symptoms appear after shipment, start from the RMA evidence collection guide and preserve the as-received condition before repair. Connect the investigation with the production records instead of replacing them with a new, untraceable narrative.