Automated Solder Inspection: A Process Guide for Consistent THT SolderJoint Quality
A solder joint is the physical and electrical connection that ties an assembly together, but consistently judging that connection at productionline speeds is not trivial. Surface finish, lead geometry, board colour, flux residue, shadows and normalprocess variations can make manual visual inspection difficult to reproduce. Automated solder inspection provides a structured way to capture images, apply validated criteria, and route relevant exceptions for review.
For throughhole technology (THT), wave soldering and selective soldering processes, the value of inspection is not only defect detection. It is the ability to establish a repeatable visual standard, retain evidence, and feed recurring defect patterns back to the source process.
What can be verified by visual solder inspection
The exact inspection list should be defined by the product’s workmanship standard and customer requirements. Typical solderjoint conditions include bridging, insufficient or excessive solder, solder balls, poor wetting, pinandlead displacement, and other processrelated surface characteristics. Not every condition can be verified by an optical camera, and not every visual difference is a defect. The correct question is: which requirements are observable at this inspection checkpoint, and what action should follow when a condition is found?
Start with a clear inspection plan: Identify the assembly side and process step to inspect. List highrisk connector, pin, and lead locations. Define acceptable examples and confirmed defects. Decide which calls can be automatically classified and which require human review. Establish containment, rework, and escalation rules.
This transforms solder inspection from informal visual spotchecks into a controlled decisionmaking process.
Why THT and wavesolder applications require dedicated consideration
THT assemblies frequently feature connectors, largeformfactor components and highdensity pins and leads that produce highly variable visual geometries. Wave and selectivesoldering processes introduce multiple sources of variation: board thermal mass, padandlead design, flux type, solder alloy composition, conveyor parameters and fixtureinduced artefacts. A single threshold or one lighting angle is seldom enough for every joint type.
The inspection method should therefore match the application. Multiangle or multicolor lighting can make slope, contour, and contrast information easier to interpret. Just as important, the program needs robust board alignment and a recipe that accounts for real component and solderjoint variation. Product changeovers, material revisions, and maintenance events should trigger a planned verification step rather than an assumption that the previous recipe still applies.
Build a defect library from real evidence
A practical solderinspection launch begins with real boards. Collect samples of acceptable joints across the full range of expected process variations and compare them against confirmed defects. Record product family, board side, joint type, image characteristics and disposition. This forms the reference dataset for recipe development and operator training.
Keep the categories useful. A long list of ambiguous alarms will slow the line without improving quality. Instead, define defect classes that correspond to clear decisions. For each class, specify whether it needs rework, engineering review, process adjustment, or no action. When a new appearance is encountered, add it to the evidence set after engineering disposition.
This discipline is especially valuable for highpincount connectors and complex lead configurations. It enables teams to investigate underlying rootcauses while avoiding the assumption that every anomalous image represents a production escape.
Connect the station to the root cause
An inline solderinspection station delivers greater value when linked to process data and clear ownership of corrective actions. Trend confirmed conditions by component, pin region, machine, product, shift, material lot, or soldering program as appropriate. For example, a localized bridging pattern may indicate the need to review fixtures, wavesolder parameters, lead design or process setup — rather than only reworking individual PCBs.
A basic feedback loop is enough to begin:
- Inspect after the defined soldering operation.
- Review only relevant exceptions using the approved visual criteria.
- Record the final disposition and repair information.
- Trend confirmed patterns.
- Assign the investigation to the responsible process owner.
- Verify the corrective action on future production samples.
This approach helps convert inspection images into evidence for continuous improvement.
Questions to ask when evaluating a solder AOI system
Use realproduction PCBs during equipment evaluation, including samples exhibiting normal process variation as well as confirmed defects. The trial shall assess imaging performance, recipe development workflow, inspection coverage, operatorreview ergonomics, line integration and data accessibility — rather than relying only on brief demonstrations using idealcondition samples.
The AIS30XHW inline THT solder automated optical inspection system is MAKERRAY’s offtheshelf solution for inline throughhole solder inspection. According to its product specification page, it adopts an RGB multispectral lighting system to expose solderjoint surface features, employs AIbased models for automatic solderjoint recognition, and supports a comprehensive range of solderrelated inspection items. The listed configurations include a 12 MP color highspeed camera, a 60 × 45 mm field of view, 15 μm resolution, and a quoted 0.23 seconds per field of view. Confirm the relevant configuration, boardsize range, and inspection performance with a representative application evaluation.
During that evaluation, ask: Which required workmanship features are visibly detectable for this joint design? How does the system handle different lead heights, connector shadows, and board finishes? Can the team easily distinguish a confirmed defect from an image that needs review? How are inspection results connected to board traceability and repair records? What is the validation process after a material, program, or product change? Does the station fit the required takt time and line interface?
Implementation rules that sustain performance
Inspection performance is maintained by process discipline. Assign ownership for recipe approval, firstarticle confirmation, alarm disposition, program changes, and maintenance verification. Retain a useful set of reference images, but review it periodically so old examples do not silently define today’s quality standard. When inspection results shift unexpectedly, investigate lighting conditions, system cleanliness, mechanical fixtures, alignment accuracy, PCB revision and process setup before broadly adjusting detection thresholds.
Automated solder inspection is not a substitute for process development or electrical test. It is a targeted visual checkpoint that helps the factory find relevant workmanship conditions earlier and respond with better evidence. Applied at the right stage and managed with clear criteria, it supports both reliable outgoing quality and a more stable soldering process.
FAQ
What is automated solder inspection?
It is a machinevision inspection process that captures solderjoint images and compares visible conditions with validated criteria to identify items that need review or action.
Is solder AOI only for wave soldering?
No. Optimal deployment depends on the assembly and manufacturing process. THT, wave soldering and selective soldering processes can all benefit from a defined opticalinspection step where visuallyverifiable features exist.
Can solder inspection prove electrical reliability?
No. Optical inspection evaluates observable workmanship conditions. Electrical verification and any required reliability testing remain separate elements of the product quality plan.
Evaluate THT solder AOI for your production line: https://www.makerrayaoi.com/en/product/detail/18