Solder Joint Visual Inspection And Acceptance Criteria

Visual inspection is the oldest and still the most common method of judging a solder joint. It is fast, it needs no capital equipment beyond a light and a lens, and it finds the defects that matter most in production. It is also the method whose results vary most between people, because the criterion is a shape that has to be recognised rather than a number that can be read.

This article describes what a visual inspection can and cannot establish, what an acceptable fillet looks like, how the common defects appear, and how the criteria and the inspectors are kept consistent.

What Visual Inspection Can And Cannot See

Visual inspection sees the surface of the joint: the shape of the fillet, the wetting angle where the solder meets the pad and the lead, the colour and the surface texture, and the presence of bridges, splashes and unmelted paste. Those observations are enough to detect the great majority of process faults, which is why the method remains the default.

It cannot see inside a joint. A void in a ball grid array, an unfilled barrel in a through hole joint, a crack that has not yet reached the surface and an intermetallic layer that is too thick are all invisible from outside. Those conditions require X-ray, a section or a destructive test, and a visual inspection that is asked to certify them is being asked the wrong question.

What An Acceptable Fillet Looks Like

A good fillet is concave and continuous, it wets the pad and the lead with a small contact angle, and it has a smooth surface without pits or cracks. The solder rises on the lead and spreads on the pad, and the transition between the two is smooth. The colour should be uniform and bright, with the slight dulling that a lead free alloy shows normal rather than a sign of a defect.

For a through hole joint the fillet has to be present on both sides, and for a surface mount joint it has to be on the pad and around the termination. The amount of solder matters as well: a joint that is under filled leaves the lead exposed and the joint weak, while an overfilled one can bury a lead so deeply that a later inspection or a rework becomes impossible.

How The Common Defects Appear

A cold joint has a dull, grainy surface and a high contact angle, and it usually means that the solder did not reach the melting point properly or that the surfaces were contaminated. A disturbed joint is similar but shows ripples or striations, because the joint was moved while it was solidifying.

Insufficient solder appears as a concave fillet that does not reach the top of the pad, and excess solder as a convex blob that hides the lead. Bridging is a connection between two adjacent joints, and it is found by looking along the row rather than at each joint in turn. Splashes and small spheres are process debris rather than joint defects, but they are a handling risk and are removed. A lifted pad or a cracked fillet appears as a line or a gap at the interface, and it is the class of defect where the appearance and the strength are most closely related.

Magnification, Lighting And Viewing Angle

Magnification is chosen for the feature being inspected. A general check of a board with standard pitch parts is done at two to four times, while a fine pitch device needs ten times or more and a joint under a low body may need the board to be tilted. The requirement is normally written into the inspection instruction, because an inspector who uses the wrong magnification will either miss defects or reject acceptable joints.

Lighting is at least as important and is often overlooked. A diffuse light from several directions reveals the shape of the fillet, while a single bright source produces a highlight that hides the wetting angle. A low angle grazing light shows the surface texture and the cracks, and a coloured background behind the joint makes the fillet outline visible. A workstation that is set up with these three conditions produces far more consistent results than one with a single lamp.

Writing The Criteria

The acceptance criteria should be written as pictures as well as words. A set of photographic standards showing acceptable, marginal and rejectable joints removes most of the argument, and the samples are retained as the reference. The written text then refers to the samples, and the standard that the product is built to is named on the drawing so that the shop knows which set applies.

The criteria differ with the alloy and with the product class. A lead free joint looks different from a tin lead joint without being defective, and a joint that is acceptable on a consumer product may be rejected on a product that has to survive thermal cycling. The alloy behaviour is described under lead free versus leaded solder, and the assembly sequence that produces the joint is set out under PCBA development process.

Training And Consistency

Two inspectors looking at the same joint should reach the same conclusion. That is achieved by training against the reference samples, by re-qualifying the inspectors periodically and by auditing their results against a second method such as X-ray or a section on a sample of boards. Where an inspector rejects far more or far fewer joints than the others, the difference is usually in the lighting or in the magnification rather than in the eyesight.

The measurement of the process itself is the number of defects found per board and the type of defect, tracked over time. A rising rate of one specific defect points to a process change, while a rising rate of every type usually points to a change of inspector or of lighting. Both are visible only if the results are recorded by defect type rather than as a single count. Where the joints are later covered, the coating that protects them is described under conformal coating and board protection.

Where The Criteria Come From

The acceptance criteria in most shops come from three sources. The first is the standard that the product is built to, which defines the classes of product and the requirements for each. The second is the customer drawing, which may tighten or relax a requirement for a specific product. The third is the internal standard of the shop, which describes how the work is done and what the inspectors are trained against. Where the three disagree, the order of precedence has to be stated, and the usual rule is that the drawing governs and the standard applies where the drawing is silent.

A criterion that is written only as a sentence is open to interpretation, and the interpretation drifts as inspectors change. The practical remedy is a reference board that carries known good, known marginal and known bad joints, kept with the inspection station and reviewed when the criteria are revised. That single artefact does more for consistency than several pages of description, and it costs almost nothing to produce from the first articles.

FAQ

Is a dull joint automatically a cold joint? No. A lead free alloy has a duller surface than a tin lead one even when it is perfectly formed, and the judgement should be based on the wetting angle and the surface texture rather than on the colour alone.

How much magnification is enough? Two to four times is adequate for standard pitch, ten times or more for fine pitch, and the figure should be stated in the inspection instruction rather than left to the inspector.

Can visual inspection replace X-ray? No. Visual inspection covers the surface and X-ray covers what is hidden, and a product that contains ball grid arrays or through hole joints with a fill requirement needs both.