Via-in-Pad in PCB Design: when is it really necessary?
As component density increases, PCB routing can become a matter of physical space even before it becomes a matter of electrical complexity. This is particularly evident under fine-pitch BGAs and other high-density packages, where the area available for traces and conventional fan-out structures becomes progressively more limited.
Via-in-pad technology addresses this limitation by placing the via directly inside the component pad. Instead of connecting the pad to an external via through the traditional dog-bone configuration, the connection can transition directly to an inner layer.
The advantage is greater routing freedom in the most densely populated areas. However, this choice introduces specific manufacturing implications: via-in-pad can reduce routing complexity by transferring part of that complexity to PCB fabrication.
When conventional fan-out becomes a limitation
In the traditional fan-out of a BGA, traces exit the pads and reach vias positioned nearby before continuing toward the inner layers. As ball pitch decreases, the space available to maintain the required trace width, clearance, and via geometries also decreases.
There is no universal pitch threshold beyond which via-in-pad becomes mandatory: the choice depends on the package, PCB technology, routing requirements, and available manufacturing capabilities. For some fine-pitch devices, however, moving the via directly into the pad makes it possible to recover space and achieve escape routing that would otherwise be very complex.
The choice of via architecture can therefore influence not only routing feasibility, but potentially also the number of layers required to complete it.
Via-in-Pad, VIPPO, and the manufacturing process
Via-in-pad and VIPPO do not mean exactly the same thing. Via-in-pad describes the geometric choice of placing a via inside the component land; it does not automatically define how that via must be manufactured.
When a flat and solderable surface is required, a VIPPO – Via-in-Pad Plated Over structure can be used. The via is filled, planarized, and subsequently plated to recreate a continuous surface over the hole. Among the different via protection structures, IPC-4761 also identifies the Type VII filled and capped configuration.
This is where the advantage gained in the layout begins to have manufacturing consequences. Filling, curing, planarization, and plating introduce additional operations and controls compared with a conventional via. Filling quality, potential voids, and surface planarity become parameters that must be controlled during fabrication.
The choice of filling material also depends on the application: a conductive fill is not automatically superior to a non-conductive epoxy. Electrical, thermal, and mechanical requirements, together with the fabricator’s qualified processes, determine the most appropriate solution.
What changes during assembly?
Via treatment becomes particularly important when the via is located within a solderable surface.
An open or inadequately treated via can allow solder to migrate into the hole during reflow, reducing the volume available for the solder joint. This issue is particularly relevant with BGAs and other bottom-terminated components, where the joints are located underneath the package and cannot be verified through simple visual inspection.
A filled, planarized, and plated-over structure can provide a more uniform mounting surface. However, it should not be prescribed indiscriminately: different packages may have different recommendations. For this reason, the semiconductor manufacturer’s guidelines and the PCB fabricator’s capabilities must be considered together.
The Trade-Off between density, cost, and lead time
Via-in-pad can simplify routing, but it does not necessarily simplify the PCB.
The additional processes required by a VIPPO structure increase fabrication complexity and can affect cost and, depending on the technology and supplier, lead time as well. They may also reduce the number of fabricators capable of supporting the required construction.
This trade-off is particularly relevant in prototyping. Recovering space may be essential for a fine-pitch BGA, but when conventional fan-out can achieve the same result without significant compromises, introducing a more complex technology may not provide a proportional benefit.
Via-in-pad can also provide benefits beyond density. A more direct transition to the inner layers can reduce interconnection length and may be useful in certain high-speed applications. In other cases, vias positioned underneath a component can contribute to heat transfer toward internal copper structures or to the opposite side of the PCB.
However, these benefits should be evaluated against specific requirements, rather than treated as reasons to adopt this technology automatically.
When is Via-in-Pad really worth using?
Fine-pitch BGA escape routing is one of the most obvious cases. AMD’s design guidelines, for example, show the use of via-in-pad for devices with a 0.5 mm pitch when the available geometry makes routing between conventional vias particularly difficult. Intel has also documented micro via-in-pad structures in its guidelines for 0.5 mm-pitch µBGAs.
This does not mean that 0.5 mm represents a universal threshold. Rather, these examples illustrate the design principle: when the available geometry becomes the real constraint, moving the via into the pad can be an effective solution.
Conversely, when the traditional dog-bone configuration easily meets dimensional, electrical, and routing requirements, maintaining a simpler construction can mean greater manufacturing flexibility and lower costs.
Conclusion
Via-in-pad is an effective tool for addressing routing and density constraints, especially under fine-pitch and high-I/O-count packages. It can free up space, shorten interconnections, and, in specific applications, provide additional electrical or thermal benefits.
But what appears to be a simple geometric choice in the layout can turn into a more specialized PCB process through filling, planarization, and plating.
For this reason, via-in-pad should be considered an engineering trade-off, not a design rule to be applied automatically. The question is not simply whether it is possible to place a via in the pad, but whether the resulting benefit is sufficient to justify the increased manufacturing complexity.