Solder Mask Dams at Fine Pitch: Registration, Width and Bridging

The strip of solder mask between two pads is a structural feature, not a cosmetic one. It keeps molten alloy from running between the pads, and at fine pitch it becomes the narrowest feature on the board, which means the whole registration budget has to be arranged around it.

What a Dam Does

Between two adjacent pads the mask forms a wall that separates them and that also defines the aperture the paste is printed into. Its width is measured at the narrowest point, and its height is the mask thickness, so the shape is an aspect ratio rather than a single dimension.

The dam has to survive stencil printing, component placement and reflow without lifting, because a dam that lifts leaves a gap where solder can wick between the pads. That is why the solder mask dam is treated as a process feature and specified on the fabrication drawing.

Minimum Width and Pitch

At a pitch of 0.5 mm with 0.3 mm pads, the dam is 0.2 mm wide, and at 0.4 mm pitch with 0.25 mm pads it falls to 0.15 mm. Below about 0.1 mm the feature is difficult to hold reliably, and most shops quote a minimum dam width in the region of 75 to 100 micrometres for photoimageable mask.

The fine pitch limit is therefore a design rule rather than a shop preference, and it is checked before the layout is released. Where the geometry does not allow a dam of the minimum width, the answer is usually a smaller pad or a different mask, not a thinner dam.

Registration and Its Budget

The dam width that remains after processing is the drawn width minus the registration error, and the error comes from the artwork, the exposure, the develop and the board’s own dimensional movement. A typical budget for a photoimageable process is plus or minus 50 micrometres, which is why a 100 micrometre dam is the practical floor.

The pad also has to keep its clearance from the edge of the opening, because a pad that is exposed at its edge can be bridged by solder. That clearance, plus the dam plus the registration, has to fit inside the pitch, and the arithmetic is worth doing explicitly rather than relying on default rules.

When the Dam Is Removed

Some designs remove the mask between pads entirely, which is done where the pitch makes a dam impossible or where the mask cannot be held reliably. The result is exposed laminate between the pads, and the solder then has to be kept off it by the paste volume and the profile rather than by a wall.

That is a legitimate choice, but it changes the failure mode from a bridge across a dam to a bridge across bare laminate, and it also exposes the laminate to flux and cleaning chemistry. The decision belongs with the assembly process rather than with the artwork alone.

Tenting and Via Coverage

Mask over a via is a tent, and a tent is a dam in one direction and a roof in the other. Tent strength depends on the mask thickness and the via diameter, and a tent over a 0.3 mm via is a more reliable feature than one over a 0.6 mm hole.

Where a via sits inside a pad, the mask is not used at all and the hole is filled and capped instead. That choice is made for the same reason a dam is used – to keep solder from wicking away from where it is wanted – and it is driven by the pitch of the device rather than by the via itself.

Mask Thickness and Aspect Ratio

Liquid photoimageable mask is normally applied at 20 to 30 micrometres over the copper, and the dam is that thickness high and its drawn width across. An aspect ratio of height to width above one is difficult to develop cleanly, which is the second reason a very narrow dam fails.

The thickness also sets the depth of the aperture that the paste is printed into, and a deeper aperture holds more paste against the pad. Where the paste volume has to be increased on a fine pitch part, the mask thickness is one of the levers, and our paste volume notes describe the others.

Bridging and Its Mask-Related Causes

A bridge between two fine pitch pads has several possible causes, and the mask is responsible for more of them than its appearance suggests. A dam that has lifted at one end, a sliver of mask left in an aperture and a run of mask that has not fully developed all change where the solder can go.

The paste volume and the profile also contribute, so the diagnosis looks at all three together. Where a bridge appears only on the boards from one panel position, the mask registration is the first thing to measure; where it appears across a whole lot, the paste or the profile is the more likely source.

Inspection and Acceptance

The dam is inspected for width, for full development and for adhesion, and the inspection is done at magnification on a sample rather than at production speed on every board. A sliver of mask in an aperture shows as a missing or reduced deposit at inspection after printing, which is often how it is found.

Adhesion is checked with the same tape test used for the rest of the mask, and a dam that fails the test is a dam that will lift during reflow. Our board quality notes describe how the visual and mechanical checks are combined so that a mask defect is caught before assembly rather than after.

Design Rules and Documentation

The design rules that matter are the minimum dam width, the minimum pad clearance to the mask opening and the minimum aperture width that the printer can fill. All three belong in the fabrication drawing rather than in a general note about the process class.

Where a design is at the limit, the drawing should say which pitch is critical, so that the shop can apply its best registration to that area. Our fabrication notes set out how those limits are written so that a released design is not renegotiated at the shop door.

FAQ

Can a dam be narrower than the registration tolerance? It can be drawn that way, and the mask will be missing on some panels, which produces a bridge rather than a narrow dam. The minimum width is set by the process capability rather than by the layout.

Is a mask-free gap between pads acceptable? It is acceptable where the paste volume and the profile keep the alloy in place, and it is a deliberate choice rather than a fallback. The exposed laminate is the price, and it has to be compatible with the cleaning chemistry.

Why does a bridge appear after a mask change? Because a mask with a different thickness changes the aperture depth and the amount of paste the aperture holds. The printer settings are re-checked after a mask change for that reason, even when the stencil is untouched.