Solder Mask Plugging Versus Tenting: Which Vias Need It
Solder mask plugging and via tenting are often treated as interchangeable ways to keep solder out of a via, but they belong to different process families and leave different surfaces behind. Tenting is a mask-only operation in which the imaged film is left intact over the drilled hole. Plugging is a separate step in which material is pushed into the barrel, either with the mask itself or with a dedicated fill compound. The choice changes surface flatness, the volume of material trapped in the barrel, the risk of blowout during reflow, and how the board behaves on the assembly line. The decision starts with the hole and with what the via is expected to do.
What Each Option Actually Does
In a tented design the solder mask is imaged so that a continuous web of film covers the via and its annular ring. Nothing enters the barrel, so the via stays as drilled and plated, and the web is supported by copper on both ends. Typical mask thickness over the copper runs 0.8 to 1.2 mil, which is thin enough that the web conforms to the annulus but has no independent strength. A tenting over vias discussion usually centres on this web and on how wide the surrounding dam has to be.
Plugging pushes material into the hole. Two variants exist. Mask plugging squeegees the liquid photoimageable mask through the barrel, then images and cures it, so the via is filled with the same dielectric used on the surface. Dedicated filling uses a resin or a thermally cured paste screen printed into the hole and often planarised afterwards. Mask plugging is a single-material route with no extra chemistry to qualify, while a resin filled via gives a flatter, more dimensionally stable result where the surface has to be built on. Both differ from a simple via plugging pass done only to stop flux from being trapped.
Why the Hole Diameter Drives the Decision
Tenting fails as a barrier when the hole is large. A web of mask film across a 20 mil barrel cannot resist the pressure of molten solder and the vapour expanding beneath it, so the film ruptures and solder wicks into the hole. As a working rule, tenting is dependable up to about 12 mil finished hole diameter and becomes progressively less so beyond that. Above roughly 16 mil the web is better described as a cover than as a seal, and the design should assume that some solder will reach the barrel.
Vias that carry current or that sit in a thermal path are usually larger than signal vias, which pushes them straight into the range where tenting is unreliable. The finished hole size, not the drill size, is what matters, because plating reduces the opening by twice the copper thickness. A drill of 14 mil that plates to 12 mil is at the boundary; the same drill in a heavy plated panel can close to 11 mil and behave acceptably. Checking the finished dimension on the fabrication drawing avoids a decision made on the wrong number.
One Side or Both Sides
Tenting one side and leaving the other open is common when the via has to be probed or when a test point is needed on the secondary side. The open end then acts as a vent during reflow, which reduces the internal pressure that ruptures the web. This arrangement works well for large thermal vias, where the goal is to keep solder on the primary side while allowing gas to escape downwards. The opening must still respect the solder mask dam rule, and the exposed copper should be treated as a solderable surface in the finish specification.
Tenting both sides gives a symmetric structure and a cleaner appearance, but it seals the barrel completely. Any moisture absorbed into the laminate or left in the hole from plating escapes only through the laminate itself, and the vapour pressure during reflow works directly on the two films. Where both ends must be covered, plugging is the more robust answer because the barrel is no longer empty. The exception is a small signal via in a thin board, where the trapped volume is negligible and a double-sided tent is entirely practical.
Via-in-Pad Changes the Rules
A via placed inside a component pad cannot be tented at all, because the pad has to remain solderable across its full area. The standard approach is to fill and planarise, then plate the pad over the fill so that the surface is flat enough for a paste deposit of consistent volume. If the fill is not planar, the stencil does not seal against the pad, paste bleeds sideways, and the joint ends up with either too little alloy or a void under the component. Flatness of the fill, not just the presence of fill, is the parameter that decides whether the assembly works.
Plating over a filled via means the copper is deposited on a dielectric plug, and that plug has to survive the plating chemistry without shrinking, outgassing or dissolving. A fill that loses volume during cure leaves a dimple that the plated copper follows, and the dimple reappears as a paste void after reflow. Requirements for this case therefore belong in the fabrication notes rather than in a verbal instruction, and they should state the fill material class, the post-fill flatness relative to the pad surface, and whether over-plating is required.
Interaction with Cleaning and Assembly
An untented, unfilled via is a passage through the board. In a no-clean process that is rarely a problem, but in a water-wash process it becomes a trap: wash chemistry enters the barrel, and the rinse may not displace it. Residue left inside the hole can later migrate to the surface and appear as a contamination defect. Plugging closes the passage and removes the question, which is one reason filled vias appear in products with tight ionic contamination limits even when the via is electrically unremarkable.
Solder balls and flux spatter also behave differently. A tented via that blows out during reflow throws molten alloy onto the surrounding surface, and the resulting spheres sit close to fine-pitch pads. A filled via has no cavity to blow out, so the failure mode disappears. Where a board has vias within a few millimetres of a fine-pitch device, that difference alone often justifies the extra process step, regardless of the electrical function of the via.
Writing the Requirement Down
The fabrication drawing should state, for each via class, whether it is tented, plugged or filled, and whether the tenting applies to one side or both. Where fill is required, the drawing should give the acceptable flatness and whether the surface will be plated over. A general note that says only “vias tented” leaves the shop to apply its own default, and that default may be tenting on one side only, which changes the venting behaviour the design assumed.
Where several via classes exist, a table keyed to the drill chart is the clearest format. Each entry names the hole size range, the treatment, and the finish on any exposed copper. The same table can be reused in the fabrication notes for later revisions, so a change in one class does not silently alter the others. Reviewing the table against the assembly drawing before release catches most of the cases where the two documents disagree.
Inspection and Acceptance
Tenting is verified visually: the film must be unbroken, and the dam width must meet the minimum. A pinhole or a thin spot at the edge of the annulus is a reject, because it is the point where the film will fail first. Inspection under magnification at a shallow angle makes the film edge visible, whereas a straight-on view hides the transition. Where the requirement is critical, cross-sectioning a coupon from the same panel confirms the mask thickness over the copper.
Filled vias are checked for voiding and for flatness. Voiding is assessed by cross-section or by X-ray on a sample basis, and the acceptance limit should be stated as a percentage of the barrel cross-section rather than as a visual impression. Flatness is measured against the surrounding pad. Both measurements belong to the first article, and both should be repeated whenever the fill material, the screen or the curing profile changes, because each of those can move the result.
FAQ
Can a tented via be used as a test point? No, unless the mask is opened at that location. A probe cannot make reliable contact through cured solder mask, and pushing through it damages the film. Where a via must serve both purposes, open the mask on the probe side and tent the opposite side, then treat the exposed copper as a solderable surface in the finish callout.
Is plugging always better than tenting? No. Plugging adds a process step, adds a material to qualify, and can introduce its own defects if the fill is not cured correctly. For small signal vias in dry-service products it buys little. It earns its cost where the hole is large, where the via sits under a pad, or where solder blowout next to fine-pitch parts would create scrap.
What happens if the mask dam is too narrow? The mask edge lifts or flakes during thermal cycling, and the opening grows. Solder can then bridge to the neighbouring feature, and the lifted film becomes a particle source. Keeping the dam at 4 to 6 mil or wider, and respecting the shop’s minimum, avoids both the lift and the bridging.
