PCB Board Thickness: Standard Values and Tolerances

Board thickness looks like the simplest parameter in a fabrication drawing and it is one of the most frequently misunderstood. A nominal figure such as 1.6 mm describes a finished board, including the copper and the plating, and the laminate that is ordered to produce it is thinner than the number suggests.

Thickness affects impedance, connector fit, mechanical stiffness and the weight of the finished product, so the value has to be chosen deliberately and specified with a tolerance rather than inherited from a previous design.

Why Thickness Matters

The electrical effect is the most direct. Trace impedance depends on the distance to the reference plane, so a change in the dielectric thickness between two layers changes the impedance of every controlled line on those layers. A board that is 10 percent thicker than the design assumes can move an impedance outside its tolerance.

The mechanical effect is equally real. A thicker board is stiffer, it resists warpage better and it accepts a press-fit connector more comfortably, while a thin board is lighter and easier to bend into a housing but more prone to movement during assembly.

The Standard Values

The common nominal thicknesses are 0.4, 0.6, 0.8, 1.0, 1.2, 1.6, 2.0, 2.4 and 3.2 mm, and 1.6 mm is the default for most rigid designs. Values between those points are available, but they usually require a specific combination of core and prepreg and are treated as a non standard build.

Thin boards, below about 0.8 mm, are frequently built from a small number of layers with a thin core. Thick boards above 2.4 mm are used for backplanes and for press-fit assemblies where the connector needs a long barrel.

Finished Versus Nominal Thickness

The finished thickness includes the base laminate, the copper on the outer layers and the plating in the holes, and it is measured across the board rather than at a single point. The inner layer copper is inside the laminate and is part of the pressed thickness, while the outer copper and the plating add to the result.

Because the outer copper and the plating vary with the design, two boards built on the same laminate can end up with different finished thicknesses. A heavy copper design can add a tenth of a millimetre or more, which matters when the thickness is specified as a hard number.

Tolerance and Measurement

A typical tolerance for a standard build is plus or minus 10 percent, and tighter tolerances are available at a cost. The measurement is taken with a micrometer at defined points, avoiding areas where the copper is dense or where the board is deformed.

Where the thickness is critical, the drawing should state the tolerance, the measurement points and whether the value applies to the finished board or to the laminate. Ambiguity here is a common source of disagreement at receiving inspection, because the two numbers can differ noticeably.

<img src="https://www.gopcba.com/wp-content/uploads/2026/08/PC3.webp" alt="Standard thickness board stackup cross section” />

Thickness and Impedance

On a controlled impedance board the dielectric thickness between the signal layer and its reference is a design input, not a consequence. The stackup is built from available core and prepreg thicknesses to reach the target, and the finished board thickness is the sum of that stack.

This is why a stackup cannot simply be scaled. Where the required impedance needs a particular dielectric thickness, the overall board thickness has to accommodate it, and if the two conflict the design has to be adjusted rather than the stack compromised.

Thickness and Connectors

Edge connectors, press-fit pins, card guides and sockets are all specified for a board thickness range. A board that is outside that range will not fit the connector properly, and forcing it damages both parts. Where the design uses a standard connector, the board thickness is effectively chosen by the connector rather than by the mechanical engineer.

Press-fit connectors are the most demanding, because the pin has a defined engagement length and the barrel has to be long enough to develop the contact. A thin board may require a longer pin or a different pin family, and the connector selection should be made together with the thickness decision rather than after it.

Special Thickness Requirements

Some designs need a non standard thickness for a specific reason: a thin board to fit a low profile housing, a thick board to support a heavy component without a separate bracket, or a stepped board where one area is milled thinner to accommodate a module.

Each of those requires a conversation with the fabricator before the design is frozen, because the available combinations of core, prepreg and copper are finite. A thickness that looks reasonable on a drawing may require a build that is expensive, slow or simply not available in the required material.

Design Rules and Specification

State the finished thickness with its tolerance, identify any area where a different thickness applies, and confirm that the stackup reaches the required impedance as well as the required overall dimension. Where a connector or a socket is involved, check the thickness range of the mating part before the stackup is finalised.

gopcb builds boards across the standard thickness range, and can propose a stackup that meets both the impedance target and the finished thickness the mechanical design requires.

Thickness, Thermal Mass and Processing

A thicker board holds more heat and takes longer to reach temperature. That affects the drilling parameters, the plating time and the desmear cycle, and it changes the way the board behaves in the reflow oven. In a thick backplane the mass of the panel means that the inner layers lag the surface, and a thermocouple placed on the surface will underestimate the peak temperature the inner layers experience.

The thermal mass also interacts with the copper distribution. A thick board with heavy copper planes needs a longer soak to reach equilibrium, and the components nearest the heavy copper see a different profile from those over a sparse area. Where a board is unusually thick or unusually heavy, the reflow profile is developed for that board rather than inherited from a thinner product with a similar circuit.

Thin Boards and Handling

Below about 0.8 mm the board becomes a handling problem before it becomes an electrical one. The panel is flexible enough to sag in the conveyor, which changes the paste deposit and the placement height, so thin boards are usually assembled in a carrier or a frame that holds the panel flat through the line.

The carrier also protects the edges from damage and provides a surface for the machine to locate against. Where a product has to use a thin board for a mechanical reason, the carrier and the panel design belong in the early planning, because they determine the panel size, the fiducial locations and the way the boards are separated afterwards.

FAQ

Is 1.6 mm always the right choice? It is the default for good reasons: it is stiff, it fits standard connectors and it is available in every material. Thin boards should be chosen for a mechanical reason rather than by preference.

Does the tolerance apply to the finished board? It should. The laminate is ordered thinner than the target so that the copper and plating bring the finished board into the specified range.

Can a board be made thicker by adding copper? To a limited extent, and only if the stackup allows. Copper is usually added for electrical reasons rather than to reach a thickness, which is better controlled through the laminate choice.

Related reading: PCB manufacturing tolerances, PCB manufacturing processes, PCB warpage and stackup balance, and multilayer PCB advantages.