Via in Pad Assembly and Void Control on Dense Boards

Placing a via inside a component pad saves routing space and shortens the current path, which is why the technique is common on dense boards and around fine pitch packages. It also creates a hole directly beneath a solder joint, and that hole competes with the joint for the paste that is printed on it. Via in pad designs fail for reasons that have nothing to do with the circuit, and most of them come down to how much solder is left after the via has taken its share.

Why Vias Are Placed in Pads

A via in the pad removes the short stub of trace that would otherwise be needed to reach the component, and it allows the decoupling capacitor or the thermal pad to sit directly over the connection to the plane below. On a high density board those savings are the difference between fitting the routing and not fitting it, and they are often the reason a design can be realised on fewer layers than would otherwise be required.

The price is that the via is now part of the joint rather than a feature somewhere else on the board. Everything that happens inside the hole, including the movement of solder and the trapping of flux vapour, happens underneath a joint that has to meet the same acceptance criteria as any other.

Paste Loss and Solder Wicking

When paste is printed over an open via, part of it drains into the hole during printing and part of the solder follows during reflow. The result is a joint that is short of solder, with a fillet that is thin or concave and a pad that is not fully covered. The loss is difficult to predict because it depends on the via diameter, the paste rheology and the print parameters, so it is normally measured on a test board before the design is released rather than estimated from a formula.

Solder wicking is the same phenomenon seen from the other side: the alloy travels down the barrel and away from the joint, sometimes reaching the opposite side of the board. Wicking is worse when the via is unfilled and when the barrel is clean and well wetted, which is precisely the condition a good plating process produces.

Via Fill and Capping Options

The usual solution is to fill the via before assembly so that it cannot accept any solder. A filled via is capped with copper and plated flat, which restores the pad as a solid surface for printing. The requirement is that the fill be complete and that the cap be thick enough to survive the assembly process without opening up again.

Where filling is not possible, a plugged or tented via is used. Tenting covers the via with solder mask, which is adequate for a via that is only used for routing but unacceptable for one that has to conduct heat or carry current, because the mask provides no electrical connection and a poor thermal path. The plating requirements behind these options are described in the guide to plating thickness.

<img src="https://www.gopcba.com/wp-content/uploads/2026/08/17aa72336fa5-1.webp" alt="X-ray image showing voiding in a solder joint over a via in pad” />

Stencil Design for Via in Pad

Printing over an unfilled via is possible if the stencil is designed for it. Reducing the aperture area over the via limits the amount of paste that can drain, and a smaller deposit is often enough for a joint that has a solid pad beneath it. The aim is to leave a controlled volume on the pad rather than to fill the hole.

A second approach is to print in two steps, with a first print that fills the via and a second that provides the joint. This costs an extra operation but it removes the competition between the hole and the joint entirely. Whatever the approach, the deposit should be measured, as described in the guide to solder paste inspection, because the useful volume is not the volume that was printed.

Void Formation and Measurement

Voids form when flux vapour or air cannot escape from the joint before the solder solidifies. A via in the pad gives the vapour an additional place to be trapped, and a filled via that is not perfectly flat gives the paste somewhere to sit unevenly. The result is a joint with a higher void content than the same joint on a solid pad.

Void content is measured by X-ray and is usually specified as a percentage of the joint area. The acceptable value depends on the application: a joint that carries current continuously is more sensitive to voids than one that only makes a signal connection, and a thermal pad has its own requirement, often expressed as a limit on the largest single void as well as on the total.

Reflow Profile Considerations

The profile affects how much vapour is generated and how easily it escapes. A slower ramp gives the flux time to activate and the volatiles time to leave the joint before the alloy melts, while a fast ramp traps them. A longer time above liquidus also allows the solder to spread and to fill the pad, which improves the appearance of the joint.

The profile has to be balanced against the rest of the assembly. Prolonging the time above liquidus to improve one joint can damage another component, and the profile should be measured on the real board with the real thermal mass rather than derived from a similar product.

Inspection and Acceptance Criteria

Inspection of a via in pad joint concentrates on two things: whether the joint is complete and whether there is a hidden void. The first is visible in an optical check, while the second needs X-ray. Both are needed because a joint that looks good can still be hollow, especially where the via is large in relation to the pad.

The criteria should be written in terms of a void percentage and a maximum single void, with a defined measurement method. Without those definitions, two inspectors can look at the same image and reach different conclusions. The techniques available for this kind of check are compared in the guide to X-ray and AOI inspection.

Reliability in Thermal Cycling

A joint over a via is subject to the same thermal expansion as any other joint, but the geometry beneath it is different. The copper of the filled via conducts heat away from the joint and constrains the pad, so the strain during a thermal cycle is concentrated at the edge of the via rather than distributed across the pad.

Where the fill is incomplete or the cap is thin, the joint can crack along that boundary. This is why the void limit for a via in pad joint is usually tighter than for a plain joint, and why thermal cycling is a standard part of the qualification for boards that use the technique.

Design Rules and Documentation

The design should state which vias must be filled, what void percentage is acceptable and what the surface finish of the capped via should be. It should also state whether the via may be tented, because the fabricator will otherwise choose the cheapest option that meets the drawing.

Traceability of these decisions matters because the assembly process depends on them. A drawing that shows a via in a pad without saying whether it is filled leaves the shop to guess, and the guess will be made on cost. The structure of the via itself and its position in the stack is covered in the guide to HDI via processing.

FAQ

Can paste be printed directly over an open via? It can, but the deposit has to be designed around the drain into the hole, and the joint will usually be short of solder unless the aperture is adjusted. The safer approach is to fill the via before assembly or to reduce the aperture deliberately and verify the result by measurement.

What void percentage is acceptable for a via in pad joint? The limit depends on the function of the joint and should be stated on the drawing. Signal joints tolerate more than current carrying joints, and thermal pads are usually specified with both a total void limit and a maximum single void limit.

Does tenting a via solve the problem? Tenting prevents solder from entering the hole, which protects the joint, but it also isolates the via electrically and thermally. It is acceptable where the via is only a routing feature and unacceptable where the via has to carry current or conduct heat away from the component.