Blind Via and Buried Via Stack Selection

A through via is the simplest way to connect two layers, and for most designs it is still the right answer. Choosing between the structures is really a question of how many lamination cycles the complete via stack requires. It is drilled once, plated in the same cycle as every other hole, and it costs nothing extra. The reason designers reach for blind and buried structures is density: when the through vias from a fine-pitch device occupy so much of every layer that there is no room left to route, the only remaining move is to stop the holes where they are needed.

Blind, Buried and Through Structures

A blind via connects an outer layer to an inner layer without passing through the whole board. Because it does not run the entire thickness, it leaves the layers below it available for routing, which is exactly the property that makes it useful in dense designs. A buried via connects two inner layers and never appears on either surface, so it consumes no surface area at all and leaves the outer layers free for components.

Both structures improve the usable space between layers, and both reduce the length of the via barrel compared with a through hole. A shorter barrel means less parasitic inductance and capacitance, which improves the electrical behaviour of the connection. The trade-off is that a blind via still occupies surface area at the layer it starts from, while a buried via does not, so the two structures solve different problems.

Sequential Lamination and What It Costs

A through via is drilled into a finished stack, but a buried via has to exist before the outer layers are added. That means the board is built in stages: an inner core is drilled and plated, then laminated with additional layers, then the next set of holes is drilled and plated. Each additional lamination cycle is the cost driver, because it introduces another press operation, another drilling operation, another plating cycle and another set of registration tolerances that accumulate.

This is why the industry counts these structures in lamination cycles rather than in numbers of vias. The first sequential lamination adds one cycle and gives one blind or buried layer pair. The second adds two, the third adds three, and the cost grows roughly with the square of the cycle count. A design that uses a second-order structure everywhere is far more expensive than one that uses it only where the routing genuinely demands it.

The number of cycles also constrains the layer counts that can be combined. The details of which structures are practical at which density are described in this article on HDI board CAM methods.

Aspect Ratio Is the Real Limit

Every plated hole has an aspect ratio: the depth of the hole divided by its diameter. The electrolyte must reach the middle of the barrel and the plating must be uniform along its length, and beyond a certain ratio neither is achievable. A shallow blind via has a favourable aspect ratio and can be made with a relatively large diameter, whereas a through via in a thick board must be drilled large enough to keep the ratio within capability.

The consequence for the designer is that via diameter is not a free choice. Reducing the diameter to gain routing space on a thick board pushes the aspect ratio up, and the fabricator will either reject the design or build it with unreliable plating. The clearance rules that surround the hole are affected as well, and their dimensional limits are set out in this discussion of via to trace clearance.

Stack Selection in Practice

The practical decision usually runs in this order. Start with through vias and see whether the design routes with them. If the escape from a fine-pitch device is the limitation, add a blind via structure that connects the outer layers to the first inner signal layers, which frees the space beneath the device and typically resolves the congestion at the lowest cost.

If the routing is still blocked, place the densest interconnections in the middle of the board, where a buried via can connect two inner signal layers without touching the outer ones. This is often the more economical structure because it consumes no surface area, and it can be built in the same lamination cycle as the blind vias.

Where the density demands a third level of interconnection, accept the additional lamination cycle only for the region that requires it, and keep the rest of the board on the simpler structure. Mixed via types on one board are normal and expected; what is not normal is applying the highest-order structure uniformly across a design that only needs it in one corner.

Electrical and Thermal Side Effects

A shorter via barrel has lower inductance, which is an advantage for high-speed signals and for power delivery. The same reduction in barrel length changes the thermal path, and a blind via of small diameter carries less current and conducts less heat than a through hole of the same diameter. Where a via is used to conduct heat out of a component, the wall area of the hole is what matters, and a stack of small blind vias may conduct less heat than the single through via it replaced.

Buried vias also change the effect of any tuning, since the stub of a through via that extends past the layer it connects to is a known source of resonance at high frequency. Removing the stub by ending the via at the target layer eliminates that resonance, which is one of the reasons these structures appear in fast serial interfaces.

Planning the Stack Before the Layout

Because the via structures determine the lamination cycles, they must be decided with the fabricator before the layout is finished, not after. A designer who routes a dense board with through vias and only then discovers that blind vias are required has to restart the routing, since the escape patterns are different. The stackup itself should be agreed at the same time, using the principles described in this article on layer stackup design.

The review should confirm four things: that each via structure is a layer count and lamination order the fabricator offers, that every hole has an aspect ratio within capability, that the clearance around each hole satisfies the process limits, and that the drill sizes are consistent between the different hole types so the plating bath can treat them uniformly.

FAQ

What is the difference between a blind via and a buried via? A blind via connects an outer layer to an inner layer and does not pass through the board, so it is visible on one surface. A buried via connects two inner layers and is not visible on either surface, so it consumes no surface area but requires the board to be laminated in stages.

Why do blind and buried vias cost so much more? Because they require sequential lamination. Each additional cycle adds a press operation, a drilling operation, a plating cycle and an extra set of registration tolerances. Cost grows roughly with the square of the cycle count, so the structures should be used only where the density demands them.

Do blind vias improve signal performance? Yes, indirectly. A blind or buried via has a shorter barrel than a through via, which reduces parasitic inductance and capacitance and removes the unused stub that can resonate at high frequency. The benefit is most visible on fast serial links and on power delivery networks.