The Trade-offs Between Cost and Performance in Via-in-Pad PCB Design

I believe a more pragmatic approach is to conduct a small-batch trial run first to validate the feasibility of the manufacturing process before deciding whether to adopt a new technology on a large scale. After all, products are meant to be sold and used—not merely displayed in a laboratory. Customers do not care how advanced the technology you employed is; they only care whether the product works well and is durable.

This principle may seem simple, yet it is easily forgotten amidst the frenetic pursuit of technical specifications. I feel that a good engineer should be like a chef: knowing how to adjust the “heat” based on the specific “ingredients” at hand, rather than blindly adhering to a rigid recipe or slavishly copying the practices of high-end establishments. Sometimes, after all, a simple home-style dish is actually more popular—wouldn’t you agree?

Whenever I see people discussing design schemes involving “via-in-pad” PCBs, I always feel that everyone tends to overcomplicate the matter. In truth, in many situations, there is absolutely no need to chase after technologies that merely sound sophisticated; instead, we should return to the most fundamental design principles.

I have encountered quite a few engineers who, the moment they encounter a BGA package, immediately insist on using via-in-pad technology—as if failing to utilize this technique would make them appear unprofessional. However, the reality is that many standard applications simply do not require such elaborate measures. Often, using ordinary through-holes or making slight adjustments to the routing strategy is sufficient to resolve the issue, while simultaneously slashing manufacturing costs significantly.

Speaking of the resin filling stage, many people assume that simply filling the voids is enough; however, there are actually many nuances involved. Different resin materials possess vastly different characteristics; for instance, if the coefficient of thermal expansion doesn’t match that of the board, problems may begin to surface after the board undergoes just a few reflow soldering cycles. The most vexing situation I have encountered is uneven surface topography after filling, which leads to uneven solder distribution during subsequent soldering processes—causing the yield rate to plummet by more than ten percentage points.

The true deciding factor in whether or not to employ via-in-pad technology is often not the technology itself, but rather whether your manufacturing environment can adequately support the process. Some factories lack the necessary precision in their grinding equipment, resulting in a surface finish that resembles the pockmarked terrain of the moon. Furthermore, some manufacturers cut corners by using substandard resin, which eventually begins to shrink and deform over time.

In fact, in many modern scenarios, we can consider a compromise. For instance, we might apply the via-in-pad treatment only to the most critical signal lines, while routing the less critical traces using traditional methods. This approach ensures signal integrity without causing manufacturing costs to skyrocket to unreasonable levels. I believe the most common mistake designers make is relying too heavily on technical data sheets while overlooking the various variables inherent in actual production. No matter how flawless your schematics and layouts appear on paper, the reality at the manufacturing plant can be an entirely different story. This is particularly true when it comes to the internal structures of multi-layer PCBs; many issues simply cannot be predicted through calculation alone.

Sometimes, instead of getting bogged down in a struggle over “via-in-pad” implementation, it is more productive to shift your perspective and look for ways to optimize the overall layout. Slightly adjusting component placement or altering the routing layers can often reveal much simpler solutions. After all, the core principle of engineering design is to solve problems using the most appropriate methods—not to show off technical prowess by employing the most complex techniques available.

Regarding the PCB manufacturing stage itself, I recommend asking your potential suppliers several practical, concrete questions when making your selection. Don’t be swayed solely by the glossy descriptions in their marketing brochures; instead, ask them to provide cross-section images of similar boards they have previously produced. Ideally, you should also visit their facility in person to observe their via-filling processes firsthand. Seeing is believing—and that carries far more weight than any technical guarantee.

Finally, I want to emphasize that PCB design always requires striking a balance between performance and cost. While the via-in-pad technique certainly offers advantages in specific scenarios, it is by no means a universal panacea. Before deciding to adopt this technology, you should first ask yourself: Is there truly no simpler alternative? Does the cost-benefit ratio justify the investment? Can your supply chain reliably support this specific manufacturing process? It is never too late to make a final decision once you have thoroughly thought through these critical questions.

I now adhere to a specific principle in my design work: if a problem can be resolved using a standard through-hole via, I will never resort to a via-in-pad; similarly, if routing can be accomplished on the outer layers, I will never unnecessarily cram traces into the inner layers. This approach may sound somewhat conservative, but it has genuinely helped me avoid a great deal of unnecessary trouble.

I recently encountered an interesting situation involving a friend who designs industrial control boards. He complained that he couldn’t achieve a satisfactory yield rate during the soldering process. He was using a BGA chip with a particularly dense pin array; while the boards appeared flawless upon visual inspection after fabrication, they proved to be unstable once installed and put into operation. After he and I spent a considerable amount of time meticulously scrutinizing the boards, we discovered that the root cause of the problem lay in that most fundamental design element: the via-in-pad implementation. Many designers assume this technique exists solely to conserve board real estate—simply dropping a via directly into a solder pad—which sounds incredibly simple and efficient on paper. However, executing it in practice is an entirely different matter—especially if you fail to properly manage the critical via-filling process. You must understand one crucial point: the small pad area on a PCB is not merely a physical connection point; it serves as a primary conduit for both electrical current and thermal energy. When you drill a hole within this area—and subsequently fill it with materials such as resin—the physical characteristics of that specific region undergo a complete transformation. I have encountered numerous instances where engineers, in their pursuit of routing convenience, indiscriminately adopted a “via-in-pad” design strategy. The result? During the reflow soldering process, a mismatch between the coefficient of thermal expansion of the filling material and that of the surrounding copper foil led to minute structural deformations. While these deformations are entirely imperceptible to the naked eye, they induce uneven stress between the BGA solder balls and their corresponding pads. Initial functional testing might be passed—perhaps barely, or even without apparent issues—yet under conditions involving thermal cycling or prolonged vibration, the probability of eventual failure increases dramatically. For instance, at the peak temperature of the reflow process, the filler material may expand too rapidly and bulge slightly; upon cooling, this leaves behind an imperceptible void that becomes a latent reliability risk over the product’s operational lifespan.

Consequently, my perspective on this matter may differ slightly. I believe the critical question regarding “via-in-pad” technology is not simply “can it be used?” but rather “should it be used?” and “how should it be used?” It is not a universal panacea for resolving high-density routing challenges. All too often, when struggling to route traces out from beneath the dense array of BGA pins, we rack our brains and naturally turn to this technique, perceiving it as the optimal solution. However, perhaps we should take a step back and consider whether alternative approaches exist. For instance, could adjusting the placement of power split planes—or optimizing the layout of peripheral components—create sufficient routing clearance for critical signals, thereby obviating the need to encroach upon the integrity of the core solder pads? Specifically, one might prioritize the adoption of finer trace width and spacing rules, or utilize blind and buried via technologies to route vias away from the active solder areas; such strategies can effectively alleviate routing congestion during the early stages of the board layout process.

I tend to view “via-in-pad” as a last resort rather than a preferred solution. This is particularly true in fields where reliability requirements are exceptionally stringent—such as automotive electronics or aerospace equipment—where PCBs must operate within environments characterized by extreme temperature fluctuations and intense vibration. In such contexts, the structural integrity and connection stability of every single component are, quite literally, a matter of life and death. You certainly have the option to select suppliers with state-of-the-art capabilities—those who can utilize laser drilling and copper plating to fill vias, rendering them nearly flush with the surrounding copper surface. However, this entails a tangible increase in both cost and lead time. In some instances, this additional expense may far exceed the savings you realized by reducing the number of board layers. For instance, a standard resin-filled via might incur only a marginal cost increase; yet, achieving perfect surface planarity and electrical conductivity requires advanced processes—such as copper plating—which can potentially double both the cost and the production cycle.

Ultimately, this decision boils down to a matter of priorities and risk assessment. If your product demands such extreme optimization regarding size and weight that the use of “via-in-pad” technology becomes absolutely indispensable, then you must channel all your design and manufacturing resources toward this specific objective. This includes conducting more rigorous simulations, performing more meticulous supplier audits, and undertaking more comprehensive reliability validations. Conversely, if there remains even a modicum of design flexibility, I sincerely advise you to steer clear of this approach and preserve those precious pad areas exclusively for the soldering process itself. After all, the true artistry of PCB design often lies not in the deployment of flashy technologies, but rather in the ability to route signals from point A to point B in the most robust and reliable manner possible; ensuring that every intermediate step is executed with absolute solidity is paramount. It is akin to constructing a building: if the foundation is laid firmly—even if the superstructure is relatively simple—it can withstand the test of time and weather; however, if one takes risks at critical load-bearing points merely to pursue novelty, latent defects may plague the structure throughout its entire lifespan.