Rigid-Flex PCB Design
Rigid-Flex PCBs are not very recent technology. They evolved from conventional Rigid PCBs and Flex PCBs with stiffeners, to fill the needs of aerospace and military requirements. They are lighter, more reliable, and allow much more design flexibility to the device they are installed in. But these benefits come at a price. Rigid-Flex PCBs are relatively costlier than Flex PCBs with stiffeners, and the design is more complex.
It used to be even more so when the design software wasn’t updated to deal with the pure designing of a rigid-flex. The designers used to break the complete Rigid-Flex PCB design into separate elements, which caused even more problems in the final assembly. Now, it has become relatively easier. But there are still a few things you need to consider while designing a Rigid-Flex PCB.
Rigid-Flex PCB Design Considerations
Most of the design considerations relate to the flex part, or the rigid to flex transition part. The rigid parts are designed mostly following conventional PCB techniques. Rigid-Flex designing also requires more electromechanical considerations as compared to a normal PCB.
It is also important to keep in close contact with your fabricator during the process of designing your Rigid-Flex, and understand their guidelines properly. Other considerations include:
Number of Layers
The number of layers is an important consideration in a Rigid-Flex PCB design. You should use as few flex layers as possible, ideally one or two. It’s even more important if you plan to mount components on the flex layer. There are always more layers on the rigid part than the flex part.
For most multilayer configurations, the flex layer (or layers) are sandwiched between the rigid layers. It requires two separate stack-up configurations. One with the Rigid-Flex part combined and one for the free flex part. When the flex layer count reaches three or up, IPC recommends that not all layers should be laminated and pressed together. Layers can be divided into pairs when possible, with each pair physically separate from another laminated pair.
This configuration is called the “air gap”. This configuration allows for more mechanical flexibility. For very short flexible curves though, four flex layers laminated together might be better than two air-gapped pairs. For such configurations, it is recommended that the adhesive layer is no more than 1/10th of the total thickness of the flex.
Type of Rigid-Flex
In terms of flexibility, Rigid-Flex is either static in design that only needs to be bent once while installing. Or it can be dynamic/flexible which requires constant bending. This significantly changes the overall buildup of the Rigid-Flex in terms of copper type (electroplater/rolled and annealed), allowance for layer count, and especially the minimum bend capability.
Dynamic Rigid-Flex is usually limited to one or two flex layers per bend. More layers flexing every time would undermine the overall integrity of the circuit. The minimum bend radius for strong dynamic Rigid-Flex should at least be 25 times the thickness. For static, the number of flex layers allowed depends upon the fabricator. The bend radius of static depends upon the number of layers. For a 1-Layer, the radius can be as low as 6 times the thickness, 10 times for 2-Layer Rigid-Flex. For more than 2 layers, the radius can lie somewhere between 10 to 15 times the thickness, depending on the number of layers.
The cost can be saved by keeping an even number of layers for the Rigid-Flex. Using an adhesiveness core is also a cost-cutting method since it reduces the thickness of the Flex area. It is also imperative that rigid areas in the Rigid-Flex all have the same number of layers. While it’s possible to create a Rigid-Flex with varying number of layers for a rigid part, but it really complicates the design process, makes it seriously costly, and you are still limited to just two different rigid area thicknesses.
One of the core benefits of a Rigid-Flex PCB design is it allows compactness to a device. But if the design is made without keeping the device layout in mind, it can create problems and costly repeat-overs. Ideally, a mechanical mock-up of the device or a doll cut-out should be made to give the designer the right idea about dimensions.
The normal bend radius employed is 10 times the thickness. The bend should never be too close to the rigid to flex transition area. There shouldn’t be any components or vias on the bend line. The use of strain relief fillets near the transition area helps in the gradual bend of the flex area, preventing any fractures.
Sharp bends should be avoided. The thinner the flex is the more bending flexibility your design will have. Like a pure flex circuit, it is important to go for the neutral bend axis to evenly divide the impact of outward tension and inward compression on the flex part. More than 90-degree bends tend to increase the tension and compression in outer and inner layers respectively.
Traces should be perpendicular to the bend lines. When the flex part has to go around corners, make sure the traces are in curves, not sharp 90 degree angles (or even 45-degree angles). For every two-layer flex, the traces should be staggered (offset) between the layers. It will increase mechanical fatigue on the flex part if traces are stacked on top of each other. It also helps reduce the overall copper thickness of the flex.
When a trace is entering a pad, don’t abruptly change its width. Use a teardrop formation instead. When designing power and ground layer, use hatched polygons instead of solid ones. Solid polygons won’t flex and forcing them would cause fractures in copper layers.
Keep-out Ratio and Vias
When designing a rigid-flex, it’s important to keep any vias away from the transition areas. IPC recommends 0.125”, but the common practice is 50 mils. Some vendors can maintain the keep-out distance as low as 36 mils without compromising integrity.
The reason for the keep out distance is the spread of the overlay to the rigid area for proper lamination. The adhesives used for flex overlay usually have a high coefficient of thermal expansions. If a via was drilled through them, the expansion would place a lot of stress on the via.
The number of vias should be kept to a minimum for the flex area.
Designing a Rigid-Flex is subjective to the requirement of the circuit. Though IPC 2223 does layout many guidelines for Rigid-Flex construction, the race to compaction has many manufacturers working in lower than recommended limits.
The technology used to design and fabricate a Rigid-Flex PCB is also continually changing. State-of-the-art manufacturers like us are pushing the boundaries and reducing costs of the manufacturing of Rigid-Flex PCBs. When you are working with us, you can design your Rigid-Flex with up to 20 layers. So no matter how complicated and compact the design is, you don’t need to worry about its seamless fabrication. For exceptional cases, we allow for aspect ratios as high as 16.
Our much larger than usual panel size (18”x24”) of 20”x80” allows you a much greater degree of design freedom. You don’t have to worry about fabrication limitations when designing ultra-long Rigid-Flex, and large format circuits when you are working with us. We also integrate HDI with Rigid-Flex, so if your circuit requirements go beyond the usual level of compactness and sophistication, you can count on us to deliver.