Conformal Coating Thickness Control and Verification

Conformal coating is specified as a thickness for a reason: the film has to be thick enough to exclude moisture and ionic contamination for the life of the product, and thin enough that it does not crack, trap stress or interfere with connectors. A coating that is too thin passes a visual inspection and fails in the field; one that is too thick looks generous and fails at the first thermal cycle. The number matters, and so does the method used to confirm it, because the measurement is not as simple as it appears on a board covered in components.

What Thickness Actually Protects

The coating works by increasing the surface insulation resistance path and by blocking the ingress of water and ionic species. Its effectiveness depends on continuity rather than on average thickness, but the specified thickness is used as a proxy for continuity because it is measurable. Thin areas, often at the edge of a component or over a sharp lead, are the points where a pinhole forms, and a pinhole over a high-impedance node is as damaging as no coating at all.

Thickness also determines the mechanical behaviour. A film below about 25 micrometres is too thin to bridge the sharp edges of a lead reliably, while a film above about 200 micrometres is thick enough to build stress at the interface during thermal cycling. Most specifications therefore land in a working band, commonly 25 to 130 micrometres for acrylic and urethane materials, with a separate figure for silicone where a thicker film is normal.

<img src="https://www.gopcba.com/wp-content/uploads/2020/05/steel_product5.jpg" alt="Eddy current probe measuring conformal coating thickness on an assembly” />

Coating Types and Their Working Ranges

Acrylic coatings are the easiest to apply and the easiest to remove, and they cure quickly at room temperature, which makes them the default for products that may need rework. Urethane coatings resist solvents and abrasion better and are harder to remove, which suits products that will not be reworked. Silicone coatings remain flexible over a wide temperature range and are used where thermal cycling is severe, at the cost of a thicker film and a longer cure.

Each chemistry has its own thickness range and its own cure mechanism. A solvent-borne acrylic loses most of its volume as the solvent evaporates, so the wet film has to be several times thicker than the finished film. A 100 % solids urethane keeps essentially all of its volume, so the wet and dry figures are close. Applying a solvent-borne material with a wet-film target taken from a 100 % solids product produces a finished film far thinner than intended, which is a common error when a product changes chemistry.

Measuring Thickness on a Finished Board

Thickness is measured either by a non-contact method on the assembled board or by a contact method on a coupon. The coupon approach uses a flat test panel coated at the same time as the product, measured with a micrometer or an eddy-current gauge, and it gives an accurate number for the process rather than for a particular point. Its weakness is that a flat coupon coats differently from a board populated with components, so it establishes that the process is in the band rather than that a given board is covered.

Non-contact measurement on the board uses eddy current over a metal surface, ultrasonic reflection through the film, or an optical method that measures the step at a masked edge. All of them need a reference: the dielectric constant of the coating, the speed of sound in it, or a clean step to measure against. For this reason the practical arrangement is to keep a coated step or a designated measurement pad on each panel, and to measure there. Where the specification is critical, a cross-section of a sacrificial assembly gives the true figure over a component.

Application Methods and Their Thickness Behaviour

Spray application deposits a film that is thickest where the gun dwells and thinnest at the edges of the pattern, so thickness depends on the robot path as much as on the gun settings. A path that crosses its own track produces a double thickness, which cracks; a path with a gap produces an uncoated strip. Dip coating coats everything that enters the tank and gives good coverage over complex shapes, but it also floods connectors, and it builds thickness at the bottom edge as the board drains.

Selective application uses a robot to coat defined areas and leaves the rest untouched, which removes much of the masking work and gives a repeatable film. Its limitation is the boundary: the edge of a coated region has a taper, and components close to the boundary may receive a thinner film or none. The overlap between adjacent passes, and the offset between the coated area and the keep-out area, should be defined in the program and verified on a coated board rather than assumed from the drawing.

Masking and Keep-Out Control

Masking defines where the coating is not wanted. Connectors, test points, adjustment pots, optical surfaces and any surface that has to make electrical contact must be protected. The mask itself has a thickness, so a tape or a boot creates a step in the coating and a shadow where the material cannot reach, and the coating thickness measured next to a mask is not representative of the rest of the board.

Masking also has to survive the cure. A tape that lifts during a thermal cure allows material under it, which then cures in a place that must be clean; a boot that is a loose fit lets material in around the edges. Where the coating is cured at elevated temperature, the mask material should be rated for that temperature. The masking plan should be drawn on the assembly, not described in words, because the difference between covering a connector body and covering its pins is a drawing question.

Cure Schedule and Its Effect on the Film

The cure schedule controls the final properties of the film. An under-cured coating remains soft and tacky, retains solvent and continues to shrink after the assembly is boxed, which pulls the film away from component edges and opens the pinholes the coating was meant to prevent. An over-cured coating becomes brittle and cracks at the corners of components, particularly on a material with a high glass transition temperature.

Cure is verified on the product, not on the oven display. The common method is a solvent rub or a pencil hardness test on a coupon, combined with a thermocouple on a board that has been through the same oven. Where the coating is UV-cured, the shadowed areas under components receive no light and depend on a secondary moisture or thermal cure, which has to be specified separately. The coating application record should carry both the oven profile and the result of the coupon test for each batch.

Adhesion, Coverage and Defect Criteria

Adhesion is checked by a tape test, by a cross-hatch test, or by a scrape with a blunt instrument on a coupon. It is the property most likely to fail silently, because a coating that has not bonded looks identical to one that has until the board is thermally cycled or handled. Adhesion failures usually trace to contamination on the board before coating, particularly flux residue or solder mask residue, which is why the wash step before coating should be specified and verified rather than assumed.

Coverage criteria should be written by area. The usual form is that the coating must be continuous over a defined region, with a limited number of pinholes accepted in areas of lower risk, and with no coating permitted in keep-out areas. Bubble, crack and delamination limits belong in the same document. The coating inspection method should state the light source and the magnification, because a defect visible under ultraviolet light may be invisible under white light and the criteria depend on which is used.

Rework and Repair of Coated Assemblies

Rework of a coated assembly starts with removing the coating locally, and the removal method depends on the chemistry: acrylic can be dissolved with a solvent, urethane usually requires mechanical removal or a dedicated stripper, and silicone is removed by cutting. The area of removal has to be larger than the area to be reworked, or the soldering iron will push coating into the joint and produce a poor connection with a harmless-looking exterior.

After rework, the area has to be re-coated and the new coating has to bond to the old. The edge of the old film should be prepared so the new material overlaps it rather than butting against it, and the cure of the repair should follow the same schedule as the original where the components allow. Where a repair is made in the field with a brush-on material, the result will not match the spray film in thickness, so the repair area should be measured and recorded. The conformal coating selection notes set out which chemistries tolerate this kind of local repair and which do not.

FAQ

Can thickness be judged by eye? Not reliably. A film that looks glossy and complete can be half the specified thickness, and the difference shows only when the board is measured or when it fails a humidity test. Visual inspection confirms coverage and cleanliness; thickness needs an instrument.

Is a thicker coating always better protection? No. Beyond a point the extra thickness builds stress at the interface, cracks at component corners and makes rework impractical. The specification should state both a minimum and a maximum, and the process should be held inside the band.

What causes a coating to peel at the edges of a component? Contamination before coating, an under-cured film that shrinks afterwards, or a film that is too thick for the temperature range it sees. The three are distinguishable: contamination gives a clean interface, under-cure gives a soft film, and excessive thickness gives a crack rather than a peel.