Step Board: 5 Rules for Cavity Depth Control

A step board is a printed circuit board with a pocket milled into it so that a tall component can sit below the surface of the panel. The pocket, usually called a cavity, is created by controlled routing depth, and it lets a connector, a shield or an optical part fit inside the overall height of the assembly. Every stepped pcb therefore depends on one number above all others, and that number is the cavity depth.

The cavity also has to respect what is left underneath it. The router removes resin and some copper, and the remaining foil forms a floor that must still carry the mechanical and electrical duty the design expects. Depth that is slightly short leaves the part proud of the surface, while depth that goes slightly too far can cut the floor and scrap the panel. A cavity pcb is a tolerance problem before it is a manufacturing one.

Why a Step Board Is Used

Height is the usual reason. A product with a fixed external envelope, such as a module that slides into a rail, cannot accept a component that stands above the board. Recessing the part into a cavity, or into a stepped area at the edge, keeps the outline intact without changing the component choice.

Cavities also help with shielding. A metal can that sits inside a pocket makes better contact with the ground ring around it, and the pocket walls limit how far solder and adhesive can spread. In both cases the cavity is a functional feature and must be toleranced like one, not treated as a cosmetic detail added at the end of fabrication.

Cavity Depth and the Copper Ledge

Below the cavity floor there is normally a copper layer that has to remain continuous, or a ground plane that carries current and heat. The distance between the floor of the pocket and the top of that copper is the ledge, and it is the number that actually matters for reliability. A floor that is too thin can delaminate or crack when the board flexes during assembly or during a thermal cycle.

The design should state the minimum remaining dielectric rather than only the cavity depth, because laminate thickness varies across the panel and between lots. Specifying both lets the shop choose the material and the process so that the floor stays above the limit at every point, including the corners where milling is least even.

Milling After Lamination or Before

A cavity can be formed after the board is laminated and finished, or built before final bonding so that the layers are pressed around a plug and the pocket appears when the plug is removed. Milling after lamination is simpler and cheaper, but it exposes resin and can leave a floor of uneven thickness. Forming the cavity before bonding gives a flatter floor and a cleaner ledge.

The choice depends on the depth and on the shape. Shallow pockets with generous dimensions are usually milled after lamination. Deep cavities, or cavities with a controlled step and a strict floor, are better formed during lamination, because the router no longer has to judge its depth against a stack that has already moved. That decision belongs in CAM review, where the stack and the tool path can be checked together.

Setting Routing Depth on the Machine

Depth is set by a combination of the tool, the pressure foot and the machine zero reference. The reference is taken on the top surface of the panel, and any foreign material, mask thickness or local warpage shifts it. For that reason the surface should be probed near the cavity rather than at one fixed point on the table.

Warpage matters because a bowed panel changes the effective depth from one end of the pocket to the other. Where the panel warpage is significant, the cavity should be milled before the final finish, or the panel should be held flat on a vacuum fixture, since no plunge setting can compensate for a board that curves through the cut.

Tool Selection and Burr Formation

Routers that cut laminate cleanly have a specific geometry and a limited life. A dull tool pushes the material rather than shearing it, which raises a burr at the edge and leaves smeared resin on the floor. The burr is more than cosmetic, because it can stop a component from seating and it can trap plating solution or cleaning fluid.

Tool life should be counted in linear metres of cut rather than in hours, because a tool that is idle does not wear. Entry and exit points in the tool path should be placed away from the cavity edge so that break through does not occur where the wall has to stay square.

Cleaning and Residue After Milling

Milling produces fine dust that clings to the cavity, and it can travel into vias and onto pads that are already printed. The dust has to be removed before the board moves on, and the cleaning method has to suit the surface finish, since an abrasive brush will damage a thin gold or silver layer.

Resin smear on the floor is a separate problem. It can hide a delamination and it affects how well adhesive or underfill bonds inside the pocket. Where the pocket is exposed to soldering heat, the material may also release volatiles, so the laminate outgassing behaviour of the exposed resin should be considered before the design is released.

Dimensional Inspection of a Cavity

Depth is normally checked with a dial gauge or a height gauge against the panel surface, taking readings at several points across the floor and at each corner. A single central reading can pass a cavity whose corners are shallow, and the corners are exactly where a component body touches first.

Optical measurement is useful for the plan dimensions and for the wall angle, but it does not replace a contact gauge for depth. The inspection record should show the reference surface, the measured points and the acceptance limits, and the drawing and the drill chart should agree on which surface the depth is measured from.

Cavity Tolerance and the Component

The cavity has to fit the component, not the other way round. Where a part is bonded into a pocket, the clearance controls how much adhesive is squeezed out and how much the part can shift before cure. Where the part is held mechanically, the wall tolerance governs how much play remains after assembly.

The tolerance stack should be built from the finished cavity, not from the nominal drawing. Adding the part tolerance, the adhesive thickness and the cavity tolerance usually shows that one of them has to be tightened, and the cheapest one to tighten is normally the routing specification rather than the component itself.

Records and First Article

The first article for a stepped board should include the cavity depth, the remaining floor, the wall squareness and a photograph of the floor under magnification. Those four results catch the defects that cause later assembly problems, and they are quick to repeat when the tool is changed.

Records should also note the tool used, its life position and the machine reference, because a cavity that drifts after a tool change can be traced to the tool rather than to the design. Keeping the first article with the order means the next build starts from a known point rather than from memory.

FAQ

How deep can a cavity be milled into a PCB? The limit is set by the layers underneath rather than by the router. A practical rule is to keep two or three dielectric plies, or a continuous copper layer, below the floor. Anything thinner risks delamination during assembly and a floor that flexes under the component.

Should the cavity be milled before or after solder mask? Milling before mask keeps the mask out of the pocket, which avoids flaking and gives a cleaner ledge, but it exposes bare resin to the plating and cleaning steps. Milling after mask protects the ledge and is easier to schedule, so the choice usually follows how the floor will be protected afterwards.

Can a stepped board still be panelised? It can, but the step changes how the panel behaves in the printer and the reflow oven. Support tooling and a second look at panel stiffness are normally needed, and the cavity should be placed so that the panel rails and the vacuum channels are not interrupted.