PCB Dimensions: Units, Footprints and Enclosure Fit

A designer who knows the tool but not the numbers will still produce a board that cannot be assembled. Software facility is a means of expressing a decision about size, and the decision is where the engineering sits: a part placed at the wrong pitch, a hole drilled at the wrong diameter, or an outline drawn a millimetre too wide all end in the same place, which is a board that does not fit.

This article sets out the dimensional thinking that a layout depends on, from the units the data is written in to the way tolerances accumulate between a board and the enclosure it has to live in.

Why Dimensions Are a Design Decision

The PCB dimensions on a board are the result of choices between constraints that pull in opposite directions. A footprint that is generous is easy to solder and hard to fit; an outline that is tight saves material and leaves no room for the tolerance of the moulding. Nothing in the tool will make that choice for the designer.

The consequence is that dimensions have to be decided deliberately and recorded, because they are the part of the design that cannot be recovered later. Copper can be re-routed and a part can be moved; an outline that is wrong by two millimetres means a new panel, a new stencil and a new enclosure.

Units and the Cost of Mixing Them

The industry works in two unit systems at once, and both are legitimate. Metric units are the natural language of mechanical drawings, board thickness and connector pitches, while thousandths of an inch remain common for trace widths, clearances and drill tables, because so much of the process heritage is imperial.

The difficulty is that the two are close enough to be confused and far enough apart to matter. A one millimetre grid and a forty mil grid are not the same, and a footprint entered in the wrong units is smaller than intended by a factor that the eye will not catch at a normal zoom. The safe practice is to fix one system for the board origin and the placement grid, convert deliberately at the boundaries, and never leave the units field implicit in the data. Which of the two systems a given fabricator prefers is part of the wider conversation between design and fabrication.

Footprint Dimensions and the Solder Joint

A footprint is a drawing of pads, and each pad has to be long enough to form a fillet and wide enough to survive the placement tolerance of the machine that will put the part on it. That is why a footprint is not simply the land pattern of the component data sheet scaled to size, but the component body plus a deliberate allowance on each axis.

The allowance is not the same in both directions. Extra length beyond the lead gives solder somewhere to climb and absorbs placement error along the axis of the lead, while extra width beyond the lead only adds the risk of bridging. A pad dimensioned without that distinction produces either a weak joint or a short. The standards for pad geometry exist precisely to keep the two allowances separate.

Enclosure-Driven Dimensioning

When a board is designed to fit an existing case, the design starts with the moulding rather than with the circuit. The enclosure supplies the maximum outline, the position of the mounting bosses, the height available above and below the board, and the location of every opening that a connector, a switch or a display has to line up with.

Those dimensions have to be measured rather than assumed, and the measurement has to be made on the part that will actually be used, not on a drawing that may describe a different revision. Where the board has to be assembled into a frame or slid into a guide, the first pad at the lower left corner is often the better origin, because the assembly tooling references the board by its features rather than by a theoretical corner.

Free Dimensioning and the Standard Edge

Where no enclosure constrains the shape, the outline is chosen by the designer, and the sensible starting point is an existing standard. Cutting the board to the next size down in a familiar series keeps it inside a panel that the fabricator already runs, which affects both price and lead time.

Free dimensioning is not the same as arbitrary dimensioning. The outline still needs corners with a radius, mounting holes at positions the assembly fixture can use, and clear space at the edge where the panel rails will be. The rules for the outline and its mounting features apply whether or not a case exists.

Thickness, Holes and Mechanical Fit

Board thickness is a dimensional decision with electrical consequences, because it sets the layer-to-layer spacing that the impedance calculations depend on and the stiffness that the assembly will see. A connector specified for a one and a half millimetre board will not sit correctly on a two millimetre one, and the legs will not reach through a hole whose diameter was chosen for a different thickness.

Hole diameters follow the same logic. A hole for a component lead is the lead diameter plus a clearance that depends on the assembly method, and a hole for a screw is the screw diameter plus a clearance that depends on whether the screw is meant to locate the board or merely to clamp it. Both belong in the same dimension table as the outline.

Tolerance Stack-Up Between Board and Enclosure

The dimensions that matter most are the ones that accumulate. The position of a connector is the sum of the tolerance on the board outline, the tolerance on the placement of the pad, the tolerance on the connector body and the tolerance on the opening in the enclosure. Each may be small, but they add, and the total has to be smaller than the clearance available at the opening.

Performing that sum before the layout is released is much cheaper than discovering it during assembly. Where the total is too large, the fix is usually to give the connector more room in the opening, to locate the board from the connector rather than from the outline, or to tighten the one tolerance in the chain that is cheapest to control.

Dimensions in the Fabrication Data

The data that goes to the fabricator should carry the dimensions that cannot be inferred from the artwork: the finished thickness and its tolerance, the outline with its own tolerance, the position and size of every hole, the location of any scoring or slot, and the impedance targets with the layers they apply to.

A note that refers to a general standard is not a substitute, because the standard offers a range and the fabricator will choose within it. Recording the actual numbers turns a set of drawings into an instruction, and it gives the first-article inspection something to compare against. It is the same discipline that prototype builds for multilayer boards expect before a panel is released.

FAQ

Which unit system should a new design use? Either, provided it is used consistently. What matters is that the origin, the grid, the drill table and the drawing all agree, and that conversions are made deliberately and recorded.

How much clearance should a board have inside a case? Enough to absorb the sum of the outline tolerance, the assembly tolerance and the moulding tolerance, with a margin for the coating. Where the board is a locating feature, the clearance is taken at the mounting features instead.

Can a footprint be scaled to fit a smaller part? No. A land pattern is dimensioned around the lead and the solder fillet it has to form, so changing the scale changes the joint. A different package needs its own footprint.