Solder Mask Dam Design Guide

A solder mask dam is the strip of mask that runs between two adjacent pads. It looks like a leftover of the mask artwork, and it is one of the features that decides whether a fine pitch component can be assembled at all. A dam that is wide enough stops solder from bridging two pads; a dam that is too narrow either breaks during fabrication or is removed altogether by the shop, and the bridging returns.

What the Dam Does

The dam physically separates the two solder volumes that form on the pads during reflow. Solder that is molten will flow across any surface it can wet, and bare laminate between two pads is a path that lets the two fillets join.

The mask is not a wetting surface, so a strip of it between the pads interrupts that path. The solder stays on the pad, forms a fillet at the edge of the mask, and the joint is defined by the mask opening rather than by the copper.

The dam also protects the laminate. Without it the copper etching and the surface finish are exposed between pads, and flux residue and moisture sit on the exposed surface.

On a component with a pitch of half a millimetre or more, the dam is easy to provide. Below that the dam width falls into the range where the mask process itself becomes the limit.

Dam Width and the Process Limit

The dam width is the gap between two mask openings, and it is defined by the artwork rather than by the copper. A typical shop can hold a dam of about 0.1 mm on a standard process, and a finer process can hold less.

The limit is set by the resolution of the mask material and by the registration between the mask and the copper. A dam that is drawn at the limit will vary in width across the panel, and in the worst case two openings will merge into one.

Where the dam cannot be held, two options remain: remove the dam entirely and accept a mask defined pad, or use a different technique such as a mask defined opening with a thin dam on a tighter process.

The choice should be made with the shop, because the number that matters is the capability of the shop that will produce the panel rather than a generic figure.

Mask Registration and Pad Definition

Registration is the alignment between the mask opening and the copper pad, and it decides how much of the pad is exposed. A mask defined pad has an opening smaller than the copper, so the mask overlaps the copper and the solderable area is set by the mask.

A copper defined pad has an opening larger than the copper, so the solderable area is set by the copper and the mask only removes the risk of bridging. The second arrangement is more tolerant of registration, and it is the usual choice for a fine pitch part.

The registration tolerance also sets the dam width in production, because the openings move relative to each other as well as to the copper. A dam that is drawn at 0.1 mm can measure 0.06 mm on one panel and 0.14 mm on the next.

Where the mask thickness is high, a narrow dam is also more likely to lift or to break during the thermal cycle, because a tall thin feature is mechanically weak.

Solder Bridging and Its Causes

Solder bridging is the formation of an unintended connection between two pads during assembly. It has three common causes: too much paste, a pad that is too large for the paste volume, and a missing or broken dam.

The paste volume is set by the stencil aperture and the stencil thickness, so a stencil that is designed for a larger pitch will overprint a fine pitch part and the excess solder has to go somewhere. A dam reduces the volume that can escape sideways.

A pad that is wider than the lead also allows the solder to spread sideways toward the neighbour. The pad geometry and the dam width have to be designed together, because a generous pad and a narrow dam is a bridging combination.

A dam also fails in the other direction, by holding flux residue against the laminate. The mask edge is a place where the cleaning chemistry does not reach easily, so a board that is not cleaned properly will show a residue line along the dam that later becomes an electrochemical migration path.

The preheat profile matters as well, since a fast ramp boils the flux and throws solder. The reflow profile is a separate variable, but it is the one that is usually blamed for a fault that was designed into the mask.

Design Rules for Fine Pitch Parts

For a part with a pitch of 0.5 mm, the pad is about 0.3 mm wide and the dam is about 0.2 mm, which is comfortable for most shops. For a pitch of 0.4 mm the dam falls to about 0.1 mm, which is at the limit of a standard process.

Below 0.4 mm the dam becomes impractical on a standard process, and the usual approach is a mask defined pad with a very thin dam or with no dam at all. The paste volume is then reduced to compensate.

A no dam design is not a failure, provided the pads are copper defined and the paste volume is correct. What it loses is the tolerance for a process excursion, so the assembly window becomes narrower.

The component datasheet should be checked for a recommended land pattern, since the manufacturer has usually characterised the mask and the pad for the part. The shift causes list is a useful companion when the part is hard to place.

Fabrication and Inspection

The mask artwork should state the dam width and the pad definition, and the shop should confirm that it can hold them. A note that says solder mask between pads is not a specification, and it will be interpreted differently by each shop.

Inspection of the dam is done optically at low magnification, and a cross section is used to check the thickness and the sidewall angle. A dam with a sloped sidewall is thinner at the top and breaks more easily than a square one.

The mask material also matters, because a liquid photoimageable mask holds a narrow dam better than a screen printed one. The choice affects the cost and the resolution together, and it should be part of the fabrication notes.

On the first article, the dam should be inspected on the densest part of the board and after the reflow cycle. A dam that survives fabrication and fails at reflow is a thermal problem rather than a resolution problem, and the two are fixed in different ways.

FAQ

How narrow can a solder mask dam be? About 0.1 mm on a standard process and less on a fine line process, with the real limit set by the registration of the shop rather than by the artwork.

What happens if the dam is missing? The mask opening merges with its neighbour and the solder can bridge. The pad becomes mask defined and the assembly window becomes narrower.

Should pads be mask defined or copper defined? Copper defined is more tolerant of registration and is usual for fine pitch. Mask defined is used where the solderable area has to be controlled precisely.