Rigid-Flex PCB Manufacturing and Bend Area Design
A rigid-flex board combines a rigid printed circuit board with a flexible circuit in one construction. The rigid sections carry the components and provide the mechanical support, while the flexible sections connect between them and allow the assembly to be folded into the space available. Using one board instead of a rigid board plus a connector and a cable removes an assembly step, saves internal volume and removes the faults that come from a connector working loose.
Structure and Where It Is Used
The construction is described by the way the rigid and flexible layers are combined. Two layer rigid-flex uses one rigid layer and one flexible layer and suits simple connections and LED strips. Four layer rigid-flex, with two rigid and two flexible layers, is found in consumer products and small modules. Six layer rigid-flex, with four rigid and two flexible layers, is used in industrial control and medical equipment, and constructions of eight layers and above, with multiple rigid and flexible layers, appear in high end communications and aerospace work.
The applications follow the same pattern. Foldable phones, smart watches and wearable devices need the board to fold. Portable and in body medical instruments need a small volume and a high reliability. Automotive control units, sensor modules and lamp modules need resistance to vibration. Industrial sensors and inspection equipment need wiring in a confined space.
Where the Manufacturing Difficulty Lies
The process is considerably more involved than producing a rigid board. The rigid sections are normally built on FR4 and the flexible sections on polyimide, and the two materials expand at different rates, so that the lamination and the subsequent steps can produce delamination or warping. The bonding material and the pressing parameters are selected for the specific stack-up and the environment the product will see.
The bend area is the most critical region of the whole board, because the life of the product depends on it. The bend radius, the number of times the board will be flexed, the copper thickness and the design of the coverlay all take part, and the recommendations that follow from them are given at the design stage rather than discovered during production.
Design Rules for the Bend Area
The radius is the first rule. A radius that is too small concentrates the strain and cracks the copper, so a static bend, made once and then fixed, is recommended to use a radius of at least ten times the thickness, and a dynamic bend, which is flexed repeatedly, at least twenty times.
The copper thickness is the second. Half an ounce or one ounce is recommended in the flexing region, and the conductor is routed with a curved transition rather than a right angle, because a corner concentrates the stress that the bend is already applying. Where a wider conductor is needed for current, it is better to use several narrower conductors in parallel than one wide one, since a wide conductor resists bending and cracks more readily.
The coverlay protects the flexible section and keeps the copper surface intact. Vias and pads are kept out of the bend area so that the covering material is continuous there, and the plated through holes are placed in the rigid regions where the board does not move.
The fourth rule is the assessment of how many times the board will be flexed. A product that is folded once during assembly and a product that moves every time it is used have different requirements, and the design follows the requirement rather than the other way round.
Quality Control Through the Build
The incoming material, including the base laminate, the coverlay and the bonding sheets, is inspected by batch against its specification. The lamination is controlled with a defined temperature, pressure and time profile, and the bond between the rigid and flexible sections is verified because that interface is where a failure would begin.
The flexible regions are protected during the remaining process steps against mechanical damage and chemical contamination, which is a handling discipline as much as a process one. Samples are then flexed as part of a life test, and the finished boards receive an electrical test and a full visual inspection before shipment.
Where an impedance requirement exists, it can be met on both the rigid and the flexible sections. Because the two materials have different dielectric constants, the values are calculated separately for each, and the test report states which structure each measurement refers to.
The design group can be involved before the layout is finished, which is where the recommendations about the radius, the copper and the coverlay are most useful, and the assembled board can continue to flex PCB assembly or to SMT assembly without leaving the group. Fabrication runs under PCB manufacturing with the records held under quality management.
Capability and Lead Time
Boards from two to twelve layers are produced, with the flexible section starting at a thickness of nought point one three millimetres and a copper weight between half an ounce and two ounces according to the bending requirement. The minimum trace width and spacing in the flexible region is four thousandths of an inch, and the surface finish is normally immersion gold, with OSP and hot air levelling available where the design suits them. A four to six layer prototype is normally delivered in ten to fifteen working days, and a volume order in fifteen to twenty, with expedited routes discussed against the project.
A bend life of tens of thousands of cycles is normally achievable for a static bend, while a dynamic bend gives a range from a few thousand to tens of thousands according to the radius, the copper and the routing. The figure is an outcome of the design rather than a property of the material, which is why the recommendation is made from the customer’s flexing requirement.
Where the Cost Comes From
`r`n
Two things drive the cost of a rigid-flex board above that of an equivalent rigid one. The first is the number of process steps: the flexible material has to be prepared and covered, the rigid and flexible sections have to be laminated together and then separated where the flexible region is exposed, and each of those operations has its own tooling and its own inspection. The second is the yield, which falls as the layer count and the number of flexible regions rise, because a fault in a flexible section cannot be repaired.
`r`n
That is why reducing the number of flexing regions, keeping the layer count as low as the design allows, and keeping vias and pads out of the bend are worth doing at the design stage. Each of those decisions removes a process step or a source of loss, and unlike a change to the material, none of them reduces the capability of the product.
FAQ
How does it differ from a plain flexible board? A flexible board is flexible throughout, while a rigid-flex board has rigid regions that carry components and flexible regions that fold, which gives more mechanical support and higher reliability at a higher cost.
Can the impedance be controlled? Yes, on both kinds of section, calculated separately because the dielectric constants differ, with a report available.
What is the lead time? Ten to fifteen working days for a typical four to six layer prototype, and fifteen to twenty for volume, with an expedited route where the project needs it.
