Design Parameters for High-Quality Rigid-Flex PCB Construction
Rigid-flex PCBs comprise both rigid as well as flexible board materials, and they consist of layers of flexible substrates attached to rigid boards. With rigid-flex circuit boards, designers can have more options, while eliminating the need for expensive, unreliable wires connecting the rigid substrates.
Ensuring the integrity of the boards and successful manufacturing requires attention to detail. Rush PCB, the largest rigid-flex printed circuit board (PCB) manufacturer, manages four vital design parameters when configuring their PCBs and creating them.
The stackup for the PCB design needs an impeccable template for the finished product to work flawlessly. For this, designers must use appropriate software and create the rigid-flex layers of the PCB. Delays and issues related to PCB performance will surely result from improper communication with the fabricator, inefficient management, and flaws with the rigid-flex stackup.
Rush PCB recommends proper ECAD software and related tools for designing rigid-flex PCBs. The board design requires fine-tuning for region-specific stack up. The software must have board specific tools for altering the board outline, rather than making multiple complicated changes when designing the stackup.
Integration tools between ECAD and MCAD can save time when designing precise stackup for configuring rigid-flex boards. ECAD programs that support 3-D design is most suitable for rigid-flex design especially as it helps with verifying the placement and integrity of sensitive components and bends.
Ground Plane Management
The configuration for the rigid-flex PCB depends on whether it will be a flex-to-install or a dynamic-flex type. Flex-to-install PCBs need bending only once when the OEM places it in the product during assembly. On the other hand, a dynamic-flex PCB needs to bend and fold repeatedly whenever the device using it is active. Ground planes in a dynamic-flex PCB are under constant stress as it bends, and therefore, need more attention during design.
The key concern during the design of a dynamic-flex PCB is the integrity of the signals and that of the ground planes surrounding them. The designer must place emphasis on the reliability of the materials and substrates the PCB uses.
For instance, using solid copper ground planes in a dynamic-flex PCB may result in cracking and failure. Some designers prefer using annealed copper on the flexible parts. Others prefer reinforcement of Gold and Nickel plating on the ground planes. However, both the above increase the cost of the PCB while reducing flexibility.
Rush PCB recommends using cross-hatched polygon rather than solid copper for ground planes. Although this does increase flexibility while reducing the possibility of cracking during repeated flexing, it impacts high-speed signals negatively. The recommendation is to add solid copper return paths below the high-speed traces. Making the return paths 5-10 times wider than the signal traces helps improve signal integrity.
Designers have several concerns when the rigid-flex PCB must bend. Apart from ground and power planes in the bend area, pads for SMDs and through holes are also at risk when present in the area of the PCB that will bend. Designers often resort to anchoring pads using additional coverlay and plating for through holes to strengthen the PCB.
Rush PCB recommends not using components or vias very close to bends in a rigid-flex PCB. When designing bends, the recommendations are to:
- Reduce material thickness at bends
- Avoid bends that are too tight
- Avoid stretching of bend material
Mechanical stresses at the bend area will repeatedly stretch the outer layers while compressing the inner layers as the PCB bends. Unless the designer has properly configured the materials at the bend, the PCB is likely to fail prematurely.
Rush PCB UK recommends routing traces perpendicular to the bend during the design. In a rigid-flex design, the perpendicular routing helps reduce the stress on the traces when they bend. For double-sided flex design, the recommendation is to offset the traces. Staggering the traces on the top and the bottom layers makes the PCB stronger, enabling it to withstand repeated bending.
When bending traces, it is best to avoid a 90-degree bend, as the corners of the trace will be under greater stress than straight traces or curved traces. Therefore, using a gradual curve when bending the traces or using piece-wise linear curves reduces the chance of traces delaminating during bends.
Likewise, using tear-drop pads rather than round pads improves the reliability of rigid-flex boards significantly. For further copper adhesion to the flexible substrate, designers provide pads with anchor stubs.
For signal integrity, it is necessary the impedance of the high-speed signal path remains constant for the entire length of the trace. As the impedance may change as the track passes from the rigid to the flexible parts (and vice-versa) of the PCB, the ECAD tools should have the provision for altering the width of the trace to accommodate the change in impedance.
Rigid-Flex PCB Construction — Standard
Although manufacturers use several combinations and configurations for rigid-flex construction, some advanced constructions use odd layer count construction, varying flex layer counts, asymmetrical construction, air-gap flex layer constructions, blind and buried vias, multiple rigid area thicknesses, integrated ZIF constructions, shielded flex layers and more.
The standard method of constructing rigid-flex PCBs has the flex layers in the center of the construction with both the flex and rigid areas containing even layer counts. There is an even layer count in the thicker regions, and these are the rigid areas. The layer count may vary from four to eighteen layers or even more. The distinguishing feature of the standard design is that the construction sandwiches the flexible layers in its center.
Rigid-Flex PCB Construction — Odd Layer Counts
Among rigid board designs, layers with even counts are more common. However, designs with odd layer counts do exist, and have their own benefits. Similarly, rigid-flex boards may have an odd layer count.
For instance, it is possible to have a nine-layer rigid board along with three flexible circuit layers. Designs that necessitate two-sided shielding primarily use the odd layer construction. Stripline type impedance control mostly drives such requirements of two-sided shielding. The layers of flex consist of the ground + signal + ground construction, aimed at controlling RF and EMI. It is also possible to guide numerous interconnects between the rigid sections. This type of design offers a high amount of flexibility while reducing the cost.
The odd layer count construction is applicable to both the rigid and the flex parts of the PCB, but may also be independent of each other. For instance, it is possible that one side has an even layer count while the other end has an odd layer count, should the application demand such a design. The primary advantage provided by this configuration is an improved mechanical bend reliability due to improved flexibility, brought about by minimizing the flex area thickness.
Rush PCB recommends keeping bend reliability and constructions compliant with IPC 2223C. This will ensure all parts of the PCB have a high degree of reliability. To some extent, the odd layer-count configuration cuts down the cost of the design as it minimizes the final number of flex layers.
Rigid-Flex PCB Construction — Asymmetrical Design
The necessity for complex impedance usually drives asymmetric designs of rigid-flex PCBs. These designs require dielectric thickness to vary widely within the design. Constructions with asymmetrical design also improves reliability and manufacturability of the PCB by allowing a reduction of the aspect ratio of blind vias in the PCB.
However, asymmetric design may require fixtures to hold down the arrays during transportation for assembly. This is due to the introduction of warps and twists because of the unbalanced build of the PCB.
For instance, in an asymmetrical construction, the location of the flex layer may no longer be at the center of the design. Rather, the location of these layers may shift up towards the top or towards the bottom. In general, an asymmetric rigid-flex PCB design does not have any significant manufacturing concerns other than introduction of a few warps and twists.
Rigid-Flex PCB Construction — Varying Flex Layer Count
A varying flex layer count design is also common for rigid-flex PCB construction, where the layer count of the flex part varies between the rigid sections. As an example, if a PCB has three rigid sections, the second and third sections may have one or two flex layers joining them, with three or four layers joining the first to the second—this type of construction may have large variations in the configurations.
Rush PCB recommends using a construction with air-gaps in the flex layer for a varying flex layer count design to meet the IPC 2223C guidelines. Although the extra flex layers may run into the rigid section for routing, they may not surpass between the other sections. The reduction in the flex section, therefore, allows it greater flexibility and bend capability, which is the primary advantage of this design.
Rigid-Flex PCB Construction — Integrated ZIF Tail Construction
A ZIF tail integration within a rigid-flex construction is a common occurrence for the stackup. Such a construction eliminates the need for a flex circuit with a separate rigid section with a ZIF connector mounted on it. Therefore, the construction actually reduces the requirements of real-estate of the rigid area. The design also helps when the PCB is a high-density type with real-estate at a premium or a very thin design where it is not possible to incorporate the height of the ZIF connector.
The integrated ZIF tail construction actually improves the reliability of the PCB, as it eliminates the rigid section, the connector, and the associated interconnect points. The ZIF tail may require an additional stiffener using Polyimide, mainly to achieve the specified thickness requirements of the finger contact area of ZIF connector. If there are multiple layers in the flex part, the stiffener may not be necessary. However, more than four layers may make it difficult to achieve the ZIF thickness requirements. Rush PCB recommends limiting the integrated ZIF tail constructions to one and or two-layer flex configurations.
Rigid-Flex PCB Construction — Blind and Buried Vias
Rigid-flex PCB constructions may also utilize blind and buried vias, just as rigid circuit boards do. The necessity for using blind and buried vias arises from high-density applications, such as those for BGA and fine-pitch component mounting that may require via-in-pad designs. In case the blind vias require interconnects to the layers of the flexible circuits, it may be necessary for the designer to utilize an asymmetrical construction.
The number of sequential lamination cycles required by the design may limit the configuration. Typically, rigid-flex PCBs with multiple layers accommodate only a specific number of cycles of lamination, before the manufacturing methods and the dimensional tolerance of the materials prevent layers form registering effectively to one another. Again, just as in rigid PCBs, applications requiring via-in-pads can use capping and via fill.
Rigid-Flex PCB Construction — Air-Gap Construction
In rigid-flex PCBs with air-gap construction, manufacturers use isolated independent pairs of flex layers to substantially improve the flexibility. Rush PCB recommends constructions with two or more layers of flex to use this method, especially as using this method with four or more layers allows the design to conform to the guidelines of IPC 2223C.
Using air-gap construction precludes the use of any flexible adhesives within the rigid areas, ensuring the reliability of via structures within them. This method of construction also offers a very high degree of reliability and ensures a long-term operational life of the PCB.
Rush PCB recommends using coverlay on both sides of each flex pair, followed by a tiny air gap. All subsequent pairs follow a similar construction.
If the manufacturer constructed the design with bonding together of all layers, there would be significant issues of via reliability, and the flex section would have limited flexibility, to the extent of being almost non-flexible. Additionally, this construction would not meet IPC 2223C requirements.
Rigid-Flex PCB Construction — Multiple Rigid Area Thickness
Although a complex construction, some designs need rigid-flex PCB designs with multiple rigid areas of different thicknesses. Rush PCB recommends reviewing alternative options because of the manufacturing complexities that the design involves. In practice, manufacturers limit the construction to a maximum of two rigid area thicknesses, mainly due to the materials they require for the construction. The stackup is expensive, as the design is similar to manufacturing two boards at the same time.
In this design, one of the rigid sections may be considerably thicker than the other rigid section. Both sections may have plated hoes and vias. However, the manufacturer may have to limit the final thickness of the thinner section as there would be some commonality of materials between the two sections. For instance, the prepreg thickness on both rigid sections need to be the same. Presence of any extra core on one side would need mirroring it on the other side as well.
For instance, the manufacturer may start the construction as a four-layer stackup, encompassing layers two through five. They will run through all the manufacturing processes in sequence, leaving out creation of the outer profile. At this time, the manufacturer will have to start from the beginning again to add layers one and six, proceeding with all the manufacturing sequence once again, finally adding the outer profile. Unless the application demands following this process of multiple rigid area thickness, the process is an expensive one.
Rigid-Flex PCB Construction — Shielded Flex layers
Applications that require FR or EMI shielding use this method of construction, where manufacturers use specialized films in place of copper layers for shielding. Using these films saves on the expense of copper layers as well as allowing the flex construction to be thinner and hence improving its flexibility, along with an effective RF and EMI shielding.
Coverlays on the flex area can have selective multiple openings that expose the ground circuit. Electrically conductive adhesives on the films allow them to bond to the uncovered ground when laminated, thereby grounding the shield layers.
Combining the rigid PCB to the flex circuits technology allows adding a significant level of creativity, the overall reduction in packaging, and design integration. Rush PCB combines several specific constructions described above to create endless numbers of rigid-flex circuit configurations for their customers.