High-density interconnect (HDI) boards depend on microvias to compress routing, escape fine-pitch ball grid arrays, and reduce layer counts. However, microvias are not simply scaled-down plated through holes. They are laser-ablated structures with tighter aspect ratios, smaller capture pads, and specific fill and cap requirements. Ignoring these constraints leads to open circuits, plating voids, solder voiding, and thermal-cycle failures. Teams should therefore apply the complete Microvia PCB Design Rules when defining stackups, pad geometries, and HDI routing strategies.
A reliable microvia design starts with three areas: core geometry and tolerances, via-in-pad structures, and material or manufacturing compatibility. Each of these areas carries its own design limits, and every limit directly affects fabrication yield, assembly quality, and long-term product reliability.
Core Microvia Geometry: Aspect Ratio, Capture Pads, and Laser Limits
Microvias are generally defined as laser-drilled vias with a diameter of 150 µm or less, although many high-density designs now use 75 µm or 100 µm structures. The most important geometric rule is the aspect ratio, which is the relationship between via depth and via diameter. For microvias, the maximum practical aspect ratio is normally 1:1. A 100 µm microvia should therefore connect through no more than 100 µm of dielectric material. Some suppliers can push slightly beyond this limit, but doing so increases the risk of poor copper plating and barrel fatigue. When copper plating has to travel deep into a narrow laser-ablated hole, the sidewall coverage becomes uneven, creating thin spots that crack under thermal cycling.
The capture pad is another critical element. Because laser drilling and layer-to-layer registration both involve positional tolerances, the pad receiving a microvia must be large enough to prevent breakout. A common rule is to size the capture pad at least 100 µm to 125 µm larger than the via diameter. For example, a 100 µm microvia should use a capture pad of approximately 200 µm to 225 µm. Tighter pad sizes may save routing space, but they leave no margin for drill wander, film movement, or inner-layer misregistration. The result can be a partial connection that passes electrical test but fails later under mechanical stress or thermal expansion.
Designers should also maintain a solid copper area on each layer where the microvia lands. If a via lands near the edge of a plane shape, the laser can partially ablate the boundary and create an irregular connection. The annular ring should be continuous, and the surrounding keep-out area should keep other copper features away from the via edge. A typical keep-out for microvia pads is 75 µm to 100 µm, depending on the fabricator’s registration capability. Copper thickness also matters. Thinner copper foil, usually 9 µm to 18 µm, improves laser drilling consistency because the laser can remove copper more predictably before entering the dielectric.
Finally, designers should not treat microvias as drop-in replacements for through vias. Their small size creates higher current-density sensitivity and greater thermal stress per unit area. For power-carrying microvias, multiple parallel vias or wider diameters may be required. For signal vias, the shorter vertical transition of a microvia can actually reduce parasitic inductance, but only if the geometry remains within the laser and plating limits of the selected HDI fabricator.
Via-in-Pad, Stacked, and Staggered Microvia Structures
Fine-pitch packages often force designers to place microvias directly inside component pads. This is known as via-in-pad design. It is essential for 0.5 mm pitch BGAs, 0.4 mm pitch devices, and many high-density connectors. However, a via-in-pad cannot simply be drilled and left open. The via must be filled and then plated flat before soldering. If the via is not fully planarized, solder paste can wick into the hole, creating voiding, insufficient solder joints, or tombstoned components.
The most common fill methods are conductive silver-epoxy, non-conductive epoxy, or direct copper plating. After filling, the surface must be plated with a thin copper cap, typically 15 µm or more, to create a smooth pad that can accept solder. The finished pad should be flat enough to meet the assembly requirements of the specific component. For fine-pitch BGAs, excessive via-cap protrusion or dimpling causes open joints or stress concentration. Therefore, the design must specify filled and capped via-in-pad clearly in fabrication notes, rather than leaving the fill method to chance.
Stacked microvias are another common HDI feature. They allow a designer to connect layer 1 to layer 2 with one microvia, and then layer 2 to layer 3 with another microvia directly above the first. This approach saves space and supports complex routing, but it demands strict fabrication rules. Each lower microvia must be filled, capped, and planarized before the next microvia is laser drilled. If the lower via is not properly filled, the upper via will have an uneven base, causing thin plating or voids. Stacked structures also experience higher stress during reflow and thermal cycling because the copper column expands differently from the surrounding laminate.
Staggered microvias are generally more reliable than stacked microvias because they offset the connection points and distribute stress over a larger area. The trade-off is that staggered structures consume more routing space and can require additional routing layers. A practical rule is to use staggered microvias when board thickness, operating temperature range, or reliability requirements are demanding. Use stacked microvias only when the package pitch or layer budget makes staggering impossible. In both cases, the landing pads must be sized with enough copper for both the microvia and the connecting trace. Under 0.4 mm pitch BGAs, designers often use 75 µm microvias on small SMD pads, with the via fill and cap specified to keep the pad flat enough for assembly.
Real-world automotive camera modules, medical sensors, and high-speed transceivers frequently use a combination of via-in-pad and stacked microvias to meet both electrical performance and space constraints. The common failure mode is not the via itself, but the interface between the fill material, the copper cap, and the laminate. That interface must be controlled through clear design rules and close communication with the fabricator.
Material Selection and DFM Tolerances That Directly Affect Microvia Yield
Microvia reliability is strongly influenced by the base material. Laser-friendly materials with fine glass cloth and uniform resin distribution produce cleaner holes and more consistent sidewalls. Standard FR-4 with coarse glass bundles can cause rough or inconsistent ablation because the laser interacts differently with glass and resin. For high-density boards, low-CTE or spread-glass dielectrics are often preferable. These materials reduce dimensional movement during lamination, drilling, and thermal cycling, which helps maintain the small registration budget required by microvia pads.
Dielectric thickness must be selected alongside via diameter. If a designer needs a 100 µm microvia to connect between specific layers, the dielectric between those layers should generally be 100 µm or less. If the available prepreg or core is thicker, the aspect ratio rises above 1:1 and the plating process becomes less reliable. Conversely, using an excessively thin dielectric may create impedance control problems or reduce mechanical stiffness. The goal is to balance electrical requirements with the microvia’s depth limit.
Manufacturing tolerances also shape the design rules. Laser-drilled hole diameter typically carries a tolerance of ±15 µm to ±25 µm depending on the equipment and material. Layer-to-layer registration may add another ±25 µm to ±50 µm of variation. Designers must include these tolerances when calculating pad sizes, plane clearances, and trace spacing. A pad that works in a perfect simulation may fail in production if it relies on zero misregistration. Sufficient annular ring and keep-out area are therefore not optional; they are the difference between a repeatable process and an intermittent failure.
Solder mask design around microvias also requires attention. For filled and capped via-in-pad structures, the pad often replaces the standard solder mask defined pad. The solder mask opening should align with the final pad size, not the pre-fill via size. If the opening is too large, solder may spread into unintended areas. If it is too small, the effective solderable area shrinks and the joint weakens. For non-filled microvias, solder mask should be kept away from the via opening so that resin or air pockets do not become trapped.
Thermal reliability is another design-rule issue. Copper-filled microvias expand at a different rate than the surrounding laminate, and repeated thermal cycling can create radial cracks at the barrel-to-pad interface. Choosing a compatible fill material, maintaining a minimum cap thickness, and avoiding excessive aspect ratios all reduce this risk. For high-reliability applications in automotive, aerospace, industrial, and medical electronics, these rules become even more critical because rework is limited and field failure costs are high. Teams that review material limits, registration tolerances, and fill requirements before routing begins can avoid most microvia-related rework and produce HDI boards that remain stable through assembly and long-term use.
Baghdad-born medical doctor now based in Reykjavík, Zainab explores telehealth policy, Iraqi street-food nostalgia, and glacier-hiking safety tips. She crochets arterial diagrams for med students, plays oud covers of indie hits, and always packs cardamom pods with her stethoscope.
0 Comments