Solder Mask Dam Design Between Pads

A solder mask dam is the strip of mask ink that runs between two adjacent pads. Its purpose is to stop molten solder from flowing from one land to the next, and its width is one of the few dimensions in a board design that is decided by the capabilities of the mask process rather than by electrical requirements.

This article covers what a dam has to do, how its width is set, when the sliver is too narrow to survive, and when a dam should be removed altogether.

What A Dam Has To Do

A dam separates two lands and keeps them separate through printing, placement, reflow and any subsequent rework. It has to survive the same thermal excursion as the rest of the mask, keep its adhesion to the laminate and to the copper beneath it, and not crack or lift at the corner where it meets a pad. Where the pitch is fine, the dam is also the only thing standing between two deposits of paste that are a fraction of a millimetre apart.

The job is partly electrical and partly mechanical. Electrically, the dam raises the surface resistance between neighbouring conductors and reduces the chance of electrochemical migration across the gap in a humid environment. Mechanically, it keeps the mask from being lifted by a solder fillet that would otherwise creep across the gap. Both functions fail together when the dam is too narrow, and the failure mode is solder bridging, which is why the dimension is treated as critical. The conventions for pad shapes and clearances that surround it are set out under PCB pad design standards.

Dimensioning The Dam

The dam width is the space between two mask openings, and it is derived from the pad spacing minus the mask expansion on each side. If two pads are spaced four tenths of a millimetre apart centre to centre and each pad is two tenths wide, the copper gap between them is two tenths; if the mask opening is enlarged by half a tenth on each side, the remaining dam is one tenth of a millimetre, or about four thousandths of an inch.

That figure sits at the edge of what conventional screen printing can hold. A typical process produces a reliable dam down to about three to four thousandths of an inch, and below that the ink may not release cleanly from the screen, may slump after printing, or may pull back from the copper during cure. Photoimageable mask holds a narrower dam than screen printed ink, and the limit is quoted by the fabricator for a specific ink and thickness rather than as an industry number. A thinner ink film holds a narrower dam but offers less dielectric strength between conductors.

Registration And The Sliver Limit

A dam is only as wide as the registration of the mask allows. The mask pattern is aligned to the copper, but the laminate moves during lamination and cure, and the copper itself has an etch tolerance, so the actual position of the opening varies from the nominal one. The variation is added to the nominal dam on the narrow side, and the design has to accept the worst case rather than the average.

When the worst case reduces the dam to a sliver that the process cannot hold, two things happen. The ink may fail to clear, leaving a thin veil over the gap that later flakes and carries solder across it. Or the ink may fail to adhere at the edge, and a narrow strip lifts during reflow. Both produce unpredictable bridging that varies from panel to panel, which is far worse than a consistent defect because it escapes inspection. The mask ink behaviour that governs this release is described under solder mask ink thixotropy.

Copper Spacing Under The Mask

The mask dam is not the only barrier, and it should not be asked to do work that the copper spacing can do better. Below the ink there is bare laminate between the two conductors, and that laminate spacing is what sets the electrical clearance and the migration resistance. If the copper gap is generous, a marginal dam can still be acceptable, because even if a sliver of ink lifts, the distance between the conductors keeps solder from bridging.

Conversely, a wide dam over a very narrow copper gap gives a false sense of security. The mask is a thin film, it wears and it can be damaged in handling, and it does not provide the clearance that a printed circuit board relies on for insulation. Design rules that keep a defined copper gap at the finest pitch, independently of the mask, are the reason that fine pitch assemblies can be built at all. The interaction between copper geometry and the mask is covered under design guidelines for manufacturability.

When Dams Should Be Removed

At the finest pitches the dam becomes impossible, and the standard answer is to remove it. A mask defined pad, where the opening spans both pads without a strip between them, eliminates the sliver entirely. The trade is that solder can now bridge across the gap, so the process has to be controlled by paste volume and by the copper spacing instead. This is the normal arrangement under a fine pitch device, where the pads are on a pitch well below four tenths of a millimetre.

The decision is made between the designer and the fabricator, and it must be made before the artwork is released. If the design assumes a dam and the fabricator cannot hold it, the fabricator either widens the opening and loses the dam, or reduces the opening and shrinks the pad, which changes the assembly process. Stating the pad pitch and the required copper gap in the fabrication notes lets the fabricator choose the workable arrangement and quote it correctly.

Process Control And Inspection

Once the design is settled, the dam is controlled through the mask process. The screen tension, the squeegee pressure and the snap-off distance set how much ink is deposited and how much it spreads; the exposure and development set the opening size; the cure sets the adhesion and the final hardness. A change in any of these moves the dam width, and the first symptom is usually bridging at the finest pitch rather than a visible change in the ink.

Inspection looks for the two failure modes directly: a lifted edge along the dam, and a residue of ink in the gap between conductors. Both are visible under magnification after development and are worth checking on a coupon with the same pitch as the product, because a dam that holds on a coarse test pattern says nothing about the behaviour at the real pitch. The paste printing that follows has to respect the same geometry, and the way pads and deposits line up is treated under placement order and pad positioning.

FAQ

What is the minimum dam width? Roughly three to four thousandths of an inch for conventional screen printed ink, and narrower for photoimageable mask. The real limit depends on the ink, the film thickness and the registration capability of the specific fabricator, so it should be confirmed rather than assumed.

Can a dam be replaced by a wider copper gap? In electrical terms a wider gap helps, but the dam still has a mechanical function during assembly. Where the dam cannot be held, the copper gap and the paste volume have to be controlled tightly enough to prevent bridging without it.

Does a thicker mask give a stronger dam? Up to a point. A thicker film has more mechanical strength but releases less cleanly from the screen and is more likely to bridge between openings at fine pitch. Most processes settle on a film thickness that balances the two.