Mask Window: Preparation, Placement and Process Control

A mask window is the opening in the solder mask that exposes the copper pad, and its size decides how much bare metal the process leaves around the joint. Too small a window and solder cannot wet the full pad; too large a window and the mask edge sits far from the joint, which reduces the solder dam between adjacent pads and invites bridging on fine pitch parts.

Window geometry is normally defined by the designer and reproduced by the fabricator, and the agreement between the two is where most problems start. A drawing that calls for a nominal clearance without a tolerance leaves the shop to choose one, and the choice then varies between suppliers. This note covers the rules that keep the mask window predictable.

What the Solder Mask Window Is

The window is defined by the mask artwork, which is a negative of the openings. It is created by exposing and developing the resist, so the window edge has a slope and a thickness that depend on the resist, the exposure and the development time. A correctly formed window shows clean copper with a defined mask edge all round.

The window is not the same as the mask clearance in the design rules. The clearance is the distance the artwork keeps from the copper edge, while the window is the physical opening after processing. Etch tolerance, mask shrinkage during cure and registration error all move the window relative to the pad, and the design has to leave room for all three.

Window Size and Placement Tolerance

A typical design rule keeps the mask back from the pad edge by 50 to 100 micrometres on each side. That band covers registration error and prevents the mask from encroaching on the pad after processing. On a dense board the same figure is split between an encroachment allowance and a sliver allowance, and the two compete for the same space.

Tolerance should be stated explicitly. A drawing that gives the nominal clearance and a bilateral tolerance lets the fabricator plan the artwork and the process, while a drawing that gives only a nominal value invites a drift that appears as a reject at the assembly stage. Our notes on solder mask misregistration describe how that drift is measured.

Windows, Slivers and Solder Dams

Between two adjacent windows on a fine pitch device there is a strip of mask, and that strip is the solder dam. A dam that is too thin lifts during cure or breaks away in service, and a broken dam loses the separation that keeps two joints apart. The minimum reliable dam width depends on the mask chemistry and on the copper height beneath it.

A sliver is the opposite failure, a narrow isolated piece of mask that survives processing without proper support. Slivers lift, and a lifted sliver becomes a loose particle in the reflow oven. Both problems come from the same design decision, which is how the space between two pads is divided between window and dam. The working rules are set out in our notes on the solder mask dam and on sliver control.

Window and Solder Volume at Fine Pitch

On a fine pitch device the window size changes the paste volume that stays on the pad, because paste printed beyond the pad cannot wet back onto it. A window that is generous leaves paste on the mask, where it forms a solder bead during reflow. A window that is too tight leaves an unwettable ring of mask encroaching on the pad and reduces the effective pad area.

The practical effect is a volume change across the board that no stencil change can compensate for. Where a lot of boards show solder beads on the mask around fine pitch parts, the first thing to check is not the printer but the mask window on a bare board, measured with an optical system or on a section.

Registration Between Mask and Copper

Registration is the alignment of the mask image with the copper pattern underneath. It is set by the exposure tool, by the dimensional stability of the panel through the process and by the mask shrinkage that happens during cure. Because the resin shrinks towards the centre of the panel, the error is smallest in the middle and largest at the corners.

That pattern explains why the same drawing can produce acceptable boards from one supplier and rejects from another: the two shops use different panel sizes and different cure cycles. Where a design is sensitive to registration, the answer is a larger clearance, a smaller panel or a mask chemistry with lower shrinkage, and the defects that follow a loss of registration are described in our notes on solder mask defects.

Windows on Vias, Test Points and Gold Fingers

Not every opening in the mask is a pad window. Tented vias have no window at all, plugged vias have a window that is smaller than the via pad, and test points need a window wide enough for the probe to land without touching mask. Each of these has to be described separately, because a single global clearance rule will produce the wrong result for at least one of them.

Gold fingers and edge connectors are usually left fully open with a generous clearance, since mask on a finger interferes with the mating contact. Test points need the opposite treatment: a tight window keeps the probe clear of the mask, and a window that is too generous allows solder to wick onto the pad during reflow. The options for via treatment are set out in our notes on via tenting and mask plugging.

Checking the Window on a Bare Board

Verification is done on the bare board before assembly, and the measurement is the distance from the window edge to the pad edge at several positions across the panel. An optical measuring system reads it quickly, and a microsection shows the mask profile and the encroachment at the same time. Sampling the panel corners as well as the centre catches the shrinkage effect.

Measure the same feature on several panels from the same lot, because registration varies between panels as well as within one. A single panel that measures well and a lot that measures poorly is a common pattern, and it points at the fabrication process rather than the artwork. Any lot that falls outside the drawing tolerance should be rejected before assembly, when the cost is still low.

Documenting Window Rules for the Fabricator

The design file should state the intended clearance per side, the minimum acceptable, the minimum dam width and the required treatment for vias, test points and connector fingers. Writing these as separate notes, rather than as one global value, removes the ambiguity that leads to a supplier choosing a default.

Include the tolerance and the measurement method, and state which features are exempt. Where a design has a critical fine pitch device, mark it and give the clearance that applies there. The acceptance criteria for the finished coating are published by IPC, and quoting them in the drawing avoids a debate at incoming inspection.

FAQ

Is a larger mask window always better for solderability? Up to a point. A larger window exposes more copper and gives more room for the joint to form, and past that point it reduces the solder dam and increases the chance of bridging. The useful window is the one that leaves a defined dam and no paste on the mask.

Why do the two sides of a panel measure differently? Mask shrinkage during cure pulls the image towards the centre, so the opening on one side of the board can sit slightly differently from the other. Where the difference exceeds the drawing tolerance, the panel size or the cure cycle has to change.

Can the mask window be corrected after fabrication? No. The window is formed by the resist image during fabrication, and there is no process that restores mask edge definition on a finished board. The correction belongs in the artwork, the panel design or the fabrication process.