勛圖tv 繚 ENGINEERING GUIDE
Solder Mask Bridge Cracking or Flaking: A Stage-by-Stage Investigation
A solder-mask bridge that cracks, flakes or disappears needs an investigation of both its geometry and its process history. The visible missing strip is an observation, not a complete diagnosis. A bridge that never formed correctly and a sound bridge damaged during later handling need different corrective actions.
This page is for an observed failure. Use solder-mask bridge design when the task is to calculate nominal web width before a board is built. Keeping the two questions separate avoids treating every failure as proof that the CAD spacing was too small.
Record the condition before cleaning it away
Photograph the affected region with a scale or other dimensional reference and identify the board side, package reference and lot. Describe whether material is cracked, lifted, missing or attached as a loose fragment. Note whether copper or another underlying surface is exposed.
Keep a representative sample in its observed state where practical. Repeated cleaning or probing can change the fracture boundary and remove material that would help identify the failure. If loose debris creates an immediate concern, document the condition and follow the agreed containment procedure before proceeding with production.
Identify the first stage where it was visible
Review inspection records after mask processing, after surface finish where relevant, after assembly and after handling or cleaning. The point of discovery is not necessarily the point of origin. A marginal feature can survive early inspection and separate during a later exposure.
The diagnostic worksheet records the first observed stage and whether earlier evidence is available. It returns follow-up questions rather than a single root cause. If no earlier inspection exists, retain that uncertainty. A controlled repeat build with stage observations may be more informative than assigning blame from one finished-board photograph.
| Review item | What to establish |
|---|---|
| Stage | Identify earliest observed failure and the evidence available from prior stages. |
| Geometry | Use the released mask image and any approved manufacturing edits. |
| Failure mode | Distinguish cracking, lifting, missing material and loose debris. |
| Closure | Verify the correction through the relevant downstream exposure. |
Reconstruct the manufactured geometry
Check the actual copper and mask files used for the build, including any approved CAM changes. Compare the affected feature with the nominal design and the permitted manufacturing variation. Determine whether the remaining web was intentionally present or whether a shared opening should have removed it.
Opening growth and local distortion can change web width. Common translation of the mask relative to copper is a separate concern. Do not subtract the same registration allowance from several dimensions and conclude that a failure is inevitable. Use a clear geometry definition and a location-specific inspection result.
Review adhesion, formation and exposure as separate hypotheses
Material preparation, coating formation, development, cure, surface condition and later mechanical or thermal exposure can each be relevant. Their significance depends on the observed condition and process history. A hypothesis is useful when it predicts an observation that can distinguish it from alternatives.
For example, failures concentrated at one handling position warrant a different investigation from failures repeated at the same fine-pitch geometry across a panel. Neither pattern alone proves the mechanism. Review the controlled records with the process owner and avoid changing several unrelated parameters before establishing a baseline.
Bridge-failure evidence brief
Bridge-failure evidence brief
Use the current inputs and keep the assumptions
Start with the feature or process question you actually need to resolve. Fields are blank until you enter values or deliberately load an illustrative example. Example mode remains labeled after editing, so a sample is not silently presented as verified project data. Clear the example before preparing a project-specific record.
Use the units printed beside each input. Keep the drawing revision, measurement basis and source of any allowance in your controlled project records. A field that asks for a limit expects your agreed requirement; the example is not an IPC threshold or a statement of 勛圖tv capability. A numerical result should lead to a focused review, not replace it.
Changing any input invalidates the displayed result and disables its download until you calculate again. The text download contains the current inputs, results and assumptions. Nothing is uploaded, stored in analytics or automatically transferred into the Contact form. You may attach the file manually if it helps explain your enquiry; completing the worksheet is optional.
| Input | What to establish before using it |
|---|---|
| Review reference | Use a nonconfidential reference to connect this worksheet to the correct design or process revision. |
| First observed stage | Discovery timing narrows questions but does not establish a root cause. |
| Earlier-stage evidence | Retain affected and unaffected locations when comparing stages. |
| Observed condition | Describe what is visible without assuming a chemical mechanism. |
| Next observation to obtain | Keep sample and board revision references in controlled records. |
Look beyond the missing strip
Inspect adjacent coating and copper for additional damage or contamination. Determine whether the condition affects the intended assembly function, exposes an unwanted conductive path or leaves loose material. The product's acceptance requirements and customer agreement control disposition; a cosmetic judgment is not enough.
If repair is considered, define whether the affected region can be restored without interfering with component lands or later operations. The repair guide explains why coating restoration cannot substitute for an assessment of damaged copper or laminate. Some conditions may require replacement rather than local patching.
Verify the correction at the revealing stage
A proposed correction should be tested through the operation that previously exposed the weakness. Looking only at freshly cured mask may miss a failure that appears after soldering or cleaning. Use representative geometry and retain the unchanged comparison conditions.
Record the change, affected lot boundary and results at each relevant stage. If a geometry change is approved, update the controlled source rather than relying on a permanent undocumented manufacturing edit. If a process change is made, identify how it will be monitored on subsequent builds.
Send evidence that supports a decision
Provide the failing region, source data, build history, first-observed stage and any unaffected comparison samples. State the quantity at risk and the acceptance requirement that is in question. Include whether the board is populated and whether further inspection or sectioning is possible.
The next action should answer a defined uncertainty: geometry, formation, adhesion, exposure or damage. This is more productive than requesting a stronger mask without specifying what failed. A clear investigation brief helps the design, fabrication and assembly teams agree both containment and a repeatable correction.
Make the investigation falsifiable and traceable
Write down the observation separately from the explanation. An observation might be a lifted strip at a named location after cleaning. An explanation might involve adhesion, geometry or mechanical damage. Keeping those statements separate makes it possible to change the hypothesis when new evidence appears without losing the original record.
For each plausible mechanism, identify an observation that would support it and one that would weaken it. Then choose the next inspection or controlled comparison that best separates the alternatives. This is more efficient than collecting many photographs that all show the same final condition while leaving the process history unknown.
Preserve the sample population and its limits. Affected samples selected for failure analysis are valuable, but they do not automatically estimate the frequency in the whole lot. Inspection counts should state their counting unit and selection method. Individual locations and stage history often provide more useful clues than a pooled percentage.
Agree containment independently of the final root-cause conclusion. The team may need to hold a defined lot while the mechanism is still under investigation. Record the evidence used to set that boundary and revise it when justified. Do not turn uncertainty into either an unsupported blanket rejection or an unsupported release.
Close the issue with evidence at the relevant stage, a controlled correction and an owner for continued monitoring. If the correction changes design data, update the source. If it changes a manufacturing process, preserve the qualification and monitoring record. A repair that hides the symptom is not the same as a demonstrated preventive action.
Standards, source context and limits
Use the official references for their stated scope and the applicable controlled document for contractual requirements. A public scope page, product description or application note is not a complete acceptance specification. Material-supplier guidance applies to the named material and process; it is not evidence that every option is available for this board.
Agree the document revision, construction, inspection method and any exceptions with the responsible engineering and quality teams. The explanatory drawings and worksheet examples on this page are original planning aids, not factory records, qualification results or compliance certificates.