Flexible and Rigid PCB: Where the Difference Actually Matters

Ask ten engineers to describe the difference between a flexible circuit board and a rigid one and most will say flexibility, then stop. That answer hides the parts that actually change a design review. A flexible board is not a rigid board made of softer material. The material changes the current carrying capacity of a given trace, the way corners behave under stress, the shape the outline should take, and what happens after the tenth bend cycle instead of the first.

What a Flexible Circuit Board Is Good At

Flex circuits earn their cost in three ways. They save space, because a polyimide film carries conductors at a thickness that a rigid board cannot match. They remove connectors and cables, because a single continuous circuit can fold from one plane to another, and every connector removed is a contact that cannot oxidize and a mating step that cannot be assembled incorrectly. They allow movement, because a thin film can follow a hinge or a sliding mechanism that would crack a rigid board on the first cycle.

The material also has real electrical and thermal behavior worth remembering. Polyimide withstands soldering temperatures and stays stable chemically, and its dielectric properties are consistent enough for controlled-impedance work. The film is thin, which means heat spreads poorly in plane but escapes quickly through the thickness, so a flex circuit with generous copper handles local hot spots differently from a rigid board of the same current rating.

What a Rigid PCB Still Does Better

A rigid PCB remains the default for good reasons. It supports weight, holds heavy components without a stiffener, and can be produced in volume with predictable registration and yield. Its dimensional stability over a twenty-year service life comes from the glass reinforcement in the laminate, and its planarity makes it suitable for fine-pitch placement and automated assembly. A rigid board can also carry far more copper per layer, which matters when a design needs a power plane or a wide ground return.

The most practical difference is cost per unit area at volume. The same circuit that is economical in polyimide for a camera module becomes expensive when the board is 100 mm on a side. A useful comparison of the two flows is collected in PCB design and fabrication.

Current Carrying Capacity on Thin Copper

This is where the two families diverge most in practice. A flexible circuit board dissipates heat less effectively than a rigid board, because the polyimide is a poor conductor and there is no thick laminate to act as a heat spreader. The result is that a trace which is comfortable at one ampere on FR-4 may need to be noticeably wider on flex for the same temperature rise.

Width is not the only lever. Spacing helps as well, because two adjacent warm traces share a boundary and raise the local ambient. Shortening the path, splitting the current across layers joined by a via farm, and keeping return current directly beneath the forward path all reduce the heat generated in the first place. The calculation is the same one used for any board, but the allowable temperature rise is usually set lower, so the resulting width is larger.

Shape, Corners and Tear Control

A rigid board can absorb a sharp internal corner. A flex circuit cannot. Any sharp feature concentrates stress, and in a film the concentration turns into a tear that propagates along the conductor. The standard answer is to keep the outline rectangular wherever possible, because a rectangle nests efficiently in the panel and wastes less material, and to radius every internal corner generously. Where the outline has to narrow, blend the transition smoothly instead of stepping it.

Free edges need a margin as well. Conductors should stay a defined distance back from the cut edge, so that handling, punching, and any trimming operation has material to spare. Plated features near the edge are the first to be damaged, and a ragged edge that reaches a conductor creates a stress riser and an exposed copper surface at the same time. The same care applies to the slot and edge features described in board outline and mounting design, where the geometry of the cut interacts with the mechanical support.

<img src="https://www.gopcba.com/wp-content/uploads/2020/12/project_image_07.jpg" alt="Bend radius and strain relief layout on a flex circuit section” />

Bend Radius and Dynamic Flexing

There are two kinds of bending, and they are not equivalent. Install-to-fit, or one-time bend, happens once during assembly and then never again. Dynamic flexing happens thousands or millions of times over the product life. A tight radius that survives a one-time bend will fail quickly under repeated cycling, because each cycle adds a small amount of cumulative damage to the copper.

The governing rule is to keep the bend radius large relative to the total thickness of the flexible section, and to keep the conductors running perpendicular to the bend line rather than along it. A second layer carrying the same conductor, joined by vias outside the bend, is a simple form of strain relief that keeps the circuit alive if the outer trace cracks. Where a bend is repeated, reduce the number of layers in the flexing region, use rolled copper if available, and place the neutral axis so the conductors sit near it. A common approach is to leave the copper out of the immediate bend zone where it is not needed, or to widen the flex region so the same displacement produces a gentler curve.

Rigid-Flex: Getting Both

Most products do not choose one or the other. A rigid-flex stack combines rigid sections that carry the components with flexible sections that connect them, removing the cable and its connectors from the assembly. The rigid portion provides the mechanical support and the multiply copper that fine-pitch parts need, while the flexible portion provides the folding path.

The design discipline comes from the transition. The interface between rigid and flexible sections is the highest-stress area on the board, so conductors need to be covered, tapered, and often duplicated on separate layers with vias placed just outside the bend. Stiffeners are added where a component or a connector has to be mounted on a flexible section. As with any board, the process flow — layout, review, prototype, and volume build — is what determines whether the transitions are manufacturable, and the sequence is described in PCBA development process.

Choosing Between Them

The decision usually comes down to counting what the assembly has to do. If the circuit must move, or if it must fit into a cavity no rigid board can enter, flex earns its cost. If the circuit has to carry meaningful current, support heavy parts, or hold tight dimensional tolerance over decades, a rigid PCB is cheaper and more predictable. If the answer is genuinely both, rigid-flex is usually smaller and more reliable than a rigid board plus a cable, even though the boards themselves cost more.

What matters most is that the decision is made consciously. A flexible circuit board treated as a drop-in replacement for a rigid one will be laid out with the wrong trace widths, the wrong corners, and the wrong expectation of how it behaves when it is folded during assembly. Choosing the material first, and then designing to its rules, avoids the rework.

FAQ

Can a flexible circuit board use the same trace widths as a rigid board? No. Because the film spreads heat poorly, a flex circuit with the same current and temperature rise generally needs wider conductors and more spacing.

How sharp can an internal corner be on flex? Radius it generously. Sharp internal corners concentrate stress and start tears that propagate along the conductors.

What makes a bend survivable in a dynamic application? A large radius relative to the section thickness, conductors running across the bend rather than along it, fewer layers in the flexing zone, and a neutral axis that keeps the copper in low strain.

Why choose rigid-flex over a rigid board with a cable? Because it removes connectors from the assembly, which removes two contact interfaces, an assembly step, and a common source of intermittent failure.