Direct Answer
Via-in-pad is a routing strategy where a via is placed directly in a component pad. It is useful for fine-pitch BGA escape routing, HDI stackups, short current paths, and thermal transfer. It is risky when treated like a normal via, because the same feature must carry plating current, survive fabrication, and still behave like a flat solderable pad during assembly.
For the shorter reference-page definition and RFQ boundary, use Via-in-Pad PCB Design before this longer implementation guide.
For reliable assembly, via-in-pad usually means drilled, plated, filled, planarized, and capped. If the via is left open in a solder pad, molten solder can flow into the hole during reflow. That is why via-in-pad must be reviewed together with Demystifying BGA PCBs: An In-Depth Overview, not only as a layout trick.
Engineering Principles
The root problem is volume control. A BGA ball or QFN thermal pad needs a predictable solder volume to form a reliable joint. An open via adds an uncontrolled reservoir. During reflow, capillary action can pull solder into the via barrel. The pad may look correct in CAD, but the joint can starve in the oven.
The second problem is flatness. Fine-pitch packages require consistent pad coplanarity so solder paste printing and component collapse are repeatable. Filled and capped via-in-pad restores a planar copper surface. Without planarization, a dimple over the via can change paste deposit shape, trap flux, or increase voiding.
Process Options
| Option | Where it fits | Assembly risk | Cost impact |
|---|---|---|---|
| Dog-bone fanout | Larger BGA pitch and enough routing room | Low when spacing is adequate | Lowest |
| Tented via near pad | Non-critical escape when solder cannot enter pad | Mask tent can open during processing | Low |
| Via-in-pad, open | Prototypes only when risk is accepted | High solder wicking risk | Low fabrication cost, high assembly risk |
| Via-in-pad, filled and capped | Fine-pitch BGA, QFN thermal pads, HDI | Lowest practical risk | Higher fabrication cost |
Worked Example
Assume a BGA pitch leaves no room for a dog-bone trace and via between pads. Via-in-pad may be justified. The correct release note should not say only "vias in pads allowed." It should specify that pad vias must be filled, planarized, and copper capped. The assembly drawing should also align stencil aperture design with the final pad surface.
If the pitch allows dog-bone fanout, choose it unless the electrical path, current path, or thermal design requires the via under the pad. Via-in-pad is powerful, but it adds process steps that must be paid for and inspected.
DFM Checks Before Release
Check minimum laser or mechanical drill size, finished hole size, aspect ratio, fill material, cap plating thickness, and the fabricator's ability to inspect voids or dimples after planarization. For HDI boards, check whether stacked microvias are permitted or whether staggered microvias are safer for reliability. The Decoding High-Density Interconnect (HDI) PCB Technology article gives broader stackup context.
Via-in-pad also changes inspection strategy. Hidden joints under BGA packages cannot be judged by optical inspection, so X-ray inspection becomes important. Review AOI and X-Ray Inspection in PCB Assembly when the pad field is dense or when voiding limits are part of the quality plan.
Common Mistakes
The most common mistake is using via-in-pad in the footprint library without telling fabrication or assembly. The board may be quoted as a standard via process, then blocked in CAM or fail during SMT. Another mistake is filling only thermal vias under a large exposed pad while leaving nearby signal pad vias open. The solder will follow the path available to it.
Surface finish matters too. A flat finish such as ENIG is common for dense BGA work, while uneven pad topography creates additional print variation. Compare finish choices in ENIG vs OSP vs HASL for PCB Manufacturing.