Solder Mask Exposure And Registration Control

Solder mask is applied as a liquid or a dry film over the whole board and then patterned by light. The exposure step decides where the mask stays and where it is removed, and the registration of that pattern to the copper underneath decides whether the openings sit on their pads or wander onto the laminate beside them. It is a photo process, so it follows photographic rules: the image is only as sharp as the contact, and the result depends on how much energy reaches the film.

This article explains what the exposure step does, how registration is achieved and held, how energy and development shape the result, and how the finished mask is inspected.

What The Exposure Step Does

A photoimageable solder mask is coated over the panel, dried, and then exposed through a photographic tool that is opaque where the mask is to remain. Where the light reaches the coating, a photoinitiator starts a reaction that cross links the resin; where the tool blocks the light, the coating stays soluble. Development then washes away the unexposed material and leaves the cured mask behind, with openings at every pad.

The alternative is a dry film, which arrives on a carrier, is laminated to the panel and is exposed and developed in the same way. Liquid inks give a thinner coating that resolves finer features and flows into recesses, while dry film gives a more uniform thickness and is easier to handle on large panels. Which one is chosen affects the exposure energy and the development time, but not the logic of the step.

Registration To The Copper

The tool must be aligned to the copper pattern, not to the panel edge. Mechanical registration pins locate the panel to the machine, and the tool is positioned from the same datum, so the accumulated error is the sum of the panel edge tolerance, the pin position, the tool position and the dimensional change of the laminate through the previous processes. Etching and pressing both move the pattern, which is why the mask tool is generated from the measured or expected final copper rather than from the digital design alone.

For fine pitch work, optical registration is used instead. Fiducials on the panel are found by a camera and the machine offsets the tool or the panel to match, which removes the effect of laminate movement. Whichever method is used, the capability has to be stated as a number, typically plus or minus 50 microns or better for a standard process and tighter for high density work, and the design rules then have to leave that much clearance.

Exposure Energy And Image Quality

Exposure energy is the product of intensity and time, and it is measured in millijoules per square centimetre. Too little energy leaves the coating under cured, so the mask is softened during development and the openings come out larger than the tool, with rounded corners and a loss of fine dams. Too much energy drives the reaction sideways under the opaque areas of the tool, which closes the openings and, at the extreme, leaves a scum of partially cured mask in the pad that solder will not wet through.

The correct energy is found by exposing a step tablet and developing it, then reading which step survives. The value is checked on a schedule and after any lamp change or chemistry change, because a lamp ages and its output falls. The same test reveals whether the tool itself has lost contrast, since a tool with a thin chrome layer will allow light through the areas that should be dark.

<img src="https://www.gopcba.com/wp-content/uploads/2026/05/直流无刷电机PCBA-1.png" alt="Cross section of a cured solder dam beside a pad” />

Development And The Sidewall

Development removes the unexposed coating with a solvent or an aqueous solution, and the resulting solder dam has a sidewall whose angle is set by the balance between exposure and development. An ideal dam has a straight or slightly tapered wall, so that the pad opening measures what the tool says it measures and the dam stands up to the solder wave without breaking away.

Over development eats into the wall and narrows the dam, which reduces the insulation between a pad and its neighbour, while under development leaves residue at the foot of the wall and in the corners of the opening. Both are corrected by changing the conveyor speed, the spray pressure or the chemistry, and both are visible on a cross section. Where the ink itself is judged, the properties are described under solder mask ink thixotropy.

Dams, Tenting And Design Rules

A solder dam is the strip of mask between two adjacent pads, and its width is the difference between the pad spacing and the two openings. If the pad pitch is 0.5 mm and the openings are 0.3 mm wide, the dam is 0.2 mm minus the registration error, which is why a design that ignores the mask tolerance will produce a dam that disappears in production. Zero width dams are not manufactured; a dam either has a real dimension or it is removed from the design.

Tenting, which covers a via completely with mask, needs the same discipline, and the choice between a tented, an open and a plugged via is described under via in pad or plated through. The rules that bring the mask, the copper and the assembly process together are set out under manufacturable design guidelines.

Inspection Of The Finished Mask

The mask is inspected for registration, for openings that are undersize or closed, for skips and pinholes, and for adhesion. A registration check is usually visual against the copper, with a coupon measured on a toolmaker microscope or with an automated optical system that compares the mask image with the copper image. Adhesion is checked with tape or by a cross hatch test after cure, because a mask that passes every dimensional check and then lifts in the wave is still a failure.

Thickness is measured on the cured coating, on the pad shoulder and on the dam, since the three are not the same. A mask that is thin over a trace edge will not hold up to handling, and a mask that is thick across a fine pitch field will bridge and trap flux. The thickness comes from the coating method and the solids content of the ink, and it is confirmed rather than adjusted after the fact.

Tooling, Panels And Dimensional Movement

The mask tool is a photograph, and like any photograph it is fixed while the panel is not. Laminate moves during pressing, during drilling and during the several drying and baking steps that precede coating, and it moves differently in the warp and in the weft direction. A tool that is generated from the nominal design will fit the artwork but not the board. Shops therefore scale the tool to the expected movement, and designs that keep the same feature density in both directions are easier to scale than designs with a wide spread of pad sizes.

Panel size and position within the tool add to the same problem. Features near the edge of a large panel accumulate more movement than features near the centre, so a capability that holds at the middle of a panel may not hold at the corners. Where a product is sensitive to this, the practical remedies are to reduce the panel size, to keep critical fine pitch features away from the edges, or to use optical registration so the machine corrects for the movement instead of the tool being expected to absorb it.

FAQ

How much clearance should be left between the copper pad and the mask opening? The opening is usually made slightly larger than the pad, and the amount depends on the registration capability of the shop, typically 50 to 100 microns per side.

Why do the openings come out smaller than the tool? Because the coating has been exposed with more energy than intended, which cross links the material under the edge of the opaque area, or because development is under running.

Can solder mask be repaired after assembly? Small areas can be touched up with a cured repair ink, but the repair is a cosmetic and handling fix rather than a replacement for a properly patterned mask.