Conformal Coating for PCBA

Conformal Coating for PCBA: Materials, Process and Acceptance

Conformal coating is the last manufacturing step that decides whether a loaded board survives the environment it was designed for. Moisture, condensation, salt spray, dust and airborne chemicals do not fail a board immediately. They fail it in the field, months after it shipped. A coating that is specified correctly and applied over a properly cleaned assembly is what stands between a board and that outcome. A coating applied over contamination or flux residue simply seals the failure in.

Choosing between acrylic, polyurethane and silicone

The three chemistries are not interchangeable, and the choice is driven by the operating environment rather than by price.

Acrylic (AR) is the general-purpose option. It cures quickly at room temperature, is easy to rework, and gives good protection against moisture and moderate humidity. It is the right default for indoor industrial and consumer assemblies where rework access matters more than maximum chemical resistance.

Polyurethane (UR) trades reworkability for solvent and chemical resistance. It withstands fuels, oils and many industrial solvents better than acrylic and is the usual choice for automotive and equipment exposed to process fluids. It is harder to remove selectively, so rework planning has to happen before coating rather than after.

Silicone (SR) holds up where temperature range matters most. It resists moisture and thermal cycling and is the common choice for outdoor enclosures, power electronics and assemblies that see repeated thermal expansion. It is the most difficult of the three to rework.

Where a product sees more than one stress, the choice is a compromise and should be stated as one. A board that must survive both salt air and repeated thermal cycling is normally specified in silicone and accepts the rework penalty; a board that needs to be repairable in the field is normally specified in acrylic and accepts a lower chemical ceiling. The EV charger and inverter and energy storage assemblies we build are typical examples of the first case.

What has to happen before the coating goes on

Most coating failures are surface preparation failures. Three steps come first.

Cleaning. Flux residue, handling oils and ionic contamination are removed, and where specified the result is verified by an ionic contamination test. Coating over an unclean surface causes dewetting and voids that are invisible until the board is in service.

Masking. Connectors, test points, mating surfaces, press-fit holes, thermal pads and anything that must remain solderable or conductive are masked before application. Masking is planned from the customer assembly drawing rather than decided on the line, and the same drawing is where the DFM review confirms that the keep-out areas are defined well enough to be manufactured.

Moisture removal. Assemblies that have absorbed humidity are baked before coating, because trapped moisture will outgas during cure and produce bubbles in the film.

Application method

Coating is applied by selective spray, manual spray, brush or dip, selected per assembly. Connector-dense boards and mixed-technology assemblies are normally coated selectively so that keep-out areas are respected without hand masking every feature. Whatever the method, the acceptance question is the same: is the required coverage present at the required thickness, on the surfaces that need it, without bridging into areas that must stay clear.

Cure, inspection and acceptance

Cure follows the material manufacturer schedule. Inspection uses UV light where the material carries a fluorescent tracer, which makes coverage and voids visible, together with visual inspection to IPC-A-610 criteria and thickness measurement on the cured film.

Defects we watch for

• Dewetting and voids usually point to contamination or an incompatible surface rather than to the coating itself.
• Bubbles come from trapped moisture or solvent, or from coating applied too thickly in one pass.
• Crazing follows solvent stress or a cure mismatch between coats.
• Capillary creep into connectors is coating drawn under a connector housing by surface tension, prevented by correct masking and by controlling viscosity and application volume.

Where coating is not the answer

Coating is not a substitute for correct design. A board that runs hot needs thermal management, not an insulating film over the hot components. Boards with fine-pitch connectors under a coating layer become difficult to rework, so if the programme expects field rework, that should be decided before the coating is specified.

The standard the coating is judged against

Conformal coating is one of the few processes where the acceptance criteria are frequently left unstated until a dispute arises, because three different documents are in play. The material is qualified to IPC-CC-830, which is what the AR, UR and SR type designations in a specification refer to. The process is governed by the coating and cleanliness requirements of J-STD-001, which is where the cleanliness verification before coating belongs. The finished assembly is inspected to IPC-A-610, which is where coverage, voids and bridging are judged. A coating specification that names only a material type is incomplete: it does not say what coverage is acceptable, and it does not say whether cleanliness was verified. We state which of the three applies on the quotation, and where a customer has an internal coating specification we work to that instead and say so.

Coverage, thickness and what the customer must state

Three things have to be decided before coating rather than after. Which surfaces are coated. Full-coverage coating of a board with open connectors is rarely what the customer means, and the assembly drawing is the document that defines the keep-out areas. What thickness is required. Thickness is set by the material supplier's data sheet and by the customer's specification, and it is measured on the cured film rather than inferred from the coating weight. A build without a stated thickness is inspected to our standard coverage rather than to a customer figure. Whether a second coat is needed. Two thin coats are preferred to one thick coat, because a single heavy pass is the usual cause of bubbles and of coating that traps solvent underneath. Where a specification calls for a given thickness, the number of passes is part of the process decision, not a line choice.

Selective spray, manual spray or dip

The three methods trade coverage against cost and reworkability. Selective spray is the right route for a board with defined keep-out areas and enough volume to justify programming: it applies the film where it is wanted and leaves connectors and test points clear without hand masking every feature. Manual spray suits low volumes and unusual geometries, and depends more heavily on operator skill, which is why the masking plan matters more here than anywhere else. Dip gives the most uniform coverage over complex geometry, but it coats everything it touches, so masking has to be complete and the assembly must tolerate coating on both sides. For a mixed-technology board with a large connector, selective spray is usually the correct answer; for a densely populated board with no keep-out areas, dip is often more consistent. The method is chosen per assembly and stated on the quotation rather than left to the line.

Coating a board that was built elsewhere

Coating an assembly that was manufactured by someone else is possible, and comes with one condition that is worth stating plainly: we cannot certify cleanliness we did not control. Where the assemblies arrive with an unknown flux chemistry or an unknown storage history, the cleaning step becomes the critical one and we either verify the result by ionic contamination test or state in the acceptance documents that cleanliness was not verified. Assemblies stored in uncontrolled humidity before coating are baked first for the same reason. Where the previous process used a no-clean flux, that is not automatically a problem, but it has to be known, because no-clean residue and some coating chemistries do not wet each other properly.

Rework and repair of coated assemblies

Rework on a coated board follows a defined route rather than being attempted directly. The coating is removed locally, mechanically or with the solvent the material supplier specifies, the rework is performed, the area is cleaned again, and the coating is reapplied and cured. Where the material is acrylic, this is routine. Where it is polyurethane or silicone, removal is slower and the local repair is visible afterwards, which is why the choice of chemistry belongs at the design stage rather than at the coating stage. Test points and connectors left clear at masking make in-circuit and functional test possible without removing coating first; leaving them unstated means the film has to be opened later to reach them. The test scope that results is covered under PCBA testing.

Specify coating before the assembly drawing is approved

Coating is decided at the assembly drawing, not at the end of the line. Send the assembly drawing with the areas that must stay clear to alden@kilandu.com, together with the operating environment and any customer coating specification, and the material, masking plan and method are fixed before the build starts, and the ordering sequence that follows is set out step by step. The turnkey PCBA page lists coating among the services included in a build, and quality assurance describes how the inspection records for the coating step are held.