Excerpt: The Printed Circuit Designer’s Guide to…Flex and Rigid-Flex Fundamentals
Designing Flex Circuits for First-Pass Success, Part 1
The design process is arguably the most important part of the flex circuit procurement process. The decisions made in the design process will have a lasting impact, for better or worse, throughout the manufacturing cycle. In advance of providing important details about the actual construction of the flex circuit, it is of value to provide some sort of understanding of the expected use environment for the finished product.
The electronics industry serves several different markets that do not always share the same product acceptability or reliability expectations. For this reason, the electronics industry, through IPC and other standards organizations, has developed a classification system that specifies what is expected of products for different classes. The system of classification is not intended to be a measure of quality. Rather, quality is a matter of conformance to a set of established requirements for a product in a given application. Therefore, quality products can be created in each of the classifications within the system. It is generally accepted that there are three classes of product. These have been defined by IPC standards as follows:
- Class 1 – Consumer products and products for non-critical applications where cost is normally the primary driver.
- Class 2 – Higher-order products in terms of quality and reliability expectations, including telecommunications, computers and general industrial.
- Class 3 – High-reliability applications including military, aerospace, automotive and medical products.
By defining the class of the product being designed, the purchaser is letting the manufacturer know what added controls to apply to the manufacturing process and the level of care they will need in the inspection process to ensure that the customer gets the product that is best suited to the application.
The following are discussions on matters of high importance to achieving first-pass success in securing quality flexible circuits from a flex circuit vendor.
It is important to provide some information about the operational requirements for the flex circuit, especially if the circuit is to be used in a dynamic flexing application, such as for a disk drive read/write head assembly. The reason for this is the circuit vendor needs to provide a plan for proper layout strategy for manufacturing; a plan which accounts for the grain direction of the copper foil during the manufacturing process. This is because there is a measurable difference in terms of flexing performance between the machine and transverse directions of the copper foil.
Fabrication Specification Details
After the basic circuit design layout is completed, the next most important piece of information required is the fabrication specification. This document communicates to the fabricator all the pertinent details for the physical construction of the circuit and what is needed and expected in the final product. If this information is incomplete or inaccurate, or if a customer has requirements that cannot be reasonably met by a competent manufacturer, time will be unnecessarily lost, at a cost to both the customer and the vendor. For this reason, it is vitally important that the fabrication specifications are checked and rechecked before putting them out for bid. In the sage words of the master carpenter, “Measure twice, cut once.”
Manufacturing system operators need not only the dimensions of the part they are to manufacture, but also the tolerance for the important features of the product. With flexible circuits, this is something that must be done with thought, care, and consideration of the realities of flexible circuit materials.
With some features, design tolerances may be critical for the performance, fit, or further processing of the product (line widths, spaces, hole sizes, physical separation of features, positional accuracy, etc.). In these cases, the manufacturer can often employ methods to deal with the requirement on a localized basis. In the case of other features, the tolerance may be less critical, significantly less critical, or even non-critical. An important thing to keep in mind is that flexible materials are not as dimensionally stable as rigid materials, and while local features may be held in tight tolerance relative to each other, features from end to end may be less predictable. Given that flexible circuits are normally installed in some 3D form after assembly, the tight tolerances on planar measurements are often not necessary. If there are questions about a tolerance callout, the designer should contact the manufacturing engineer. It is always best to solve the problem before it becomes a problem.
Unclear Layer Designation (Rigid or Flex)
The purpose of a product specification is to provide clear, unambiguous instructions on the product’s construction. In the case of a multilayer circuit design, this is vitally important. The relationship of internal circuit layers relative to one another has become increasingly important in not only assuring that correct interconnections are being made, but also in product performance, especially with controlled impedance designs and signal integrity issues. Several different systems have been developed over the years to help assure that there is no uncertainty in the order of the circuit layers in the final construction. The fabricators engineering staff can provide recommendations if needed. Note the thickness and construction of each core in Figure 1.
Figure 1: Example of four-layer flex construction.
Coverlayer Requirements Not Properly Called Out or Defined
Coverlayer and cover coat are terms normally reserved for flexible circuit constructions and they are by default a defining structural element of both flex and rigid flex circuits. Coverlayers serve as a flexible solder mask of sorts, protecting the delicate circuits from damage and potential wicking of solder along circuit traces, while leaving open access to design features where interconnections are to be made to components by soldering.
It is important to determine the thickness of a coverlayer to allow for maximum flexibility when desired, and ensure you have chosen a coverlayer with a sufficient amount of adhesive on it to accommodate the copper weight. Coverlayers are also of importance in the design of areas where the circuit is to be bent either just one time, intermittently, or dynamically, millions or even billions of times over its useful life. The latter case, the dimensions and make of the flexible circuit coverlayer is critical. In dynamic flex circuits, there is need to balance the amount of flexible materials on the sides of the conductors where flexing is to occur. It is important to know and understand that there are different types of materials available for use as coverlayer materials, and that there is no single, ideal solution. These material choices include: materials that are laminated to the copper circuits using heat and pressure; materials that can be laminated and then pho toimaged, like solder mask, to define points of connection; and materials that are simply screen printed on to seal traces, while leaving open features of interest for further processing or for making interconnections.
Number of Flex Layers
The clear majority of flexible circuits have just one or two metal layers. However, an increasing number of high-performance products now require high layer counts and high density interconnect (HDI) design techniques. As layer count increases, so does the need for control in design generation to accommodate manufacturing process realities. It is also worth noting, while on the topic of layer count, that stiffness increases as a cube of thickness. That is, if one doubles its thickness, the stiffness goes up eightfold (23 = 8), and thus small increases in thickness due to increases in layer count can greatly decrease circuit flexibility. The converse is also true, of course. The following are some key concerns to be understood and addressed in the design process relative to flex layer count.
As is the case with any multilayer construction, core thickness must be provided with the assumption that copper is clad on at least one surface. The core thickness is generally understood to be the thickness of the dielectric material between the copper layers. The core material can be a simple single-sided piece of copper clad polymer, or it can be clad with copper on both sides. Many different core thicknesses are commonly available for flexible circuits, but the most common is 75 mm, typically comprised of 25 mm of base polymer (e.g., poly imide, polyester) with 25 mm of adhesive (e.g., acrylic, modified epoxy) on either side to bond copper foil to the surface of the base polymer. Thinner and thicker core materials can be procured both with and without adhesive. It is recommended that designers check with their flex vendors for both their recommendations and the availability of the chosen material.
While the discussion so far been limited to flexible circuit core material, rigid materials are employed in the fabrication of rigid-flex circuits. Of course, any of the myriad core materials used in rigid multilayer circuits are also available to make rigid-flex circuits. However, once again, it is advisable to check with the flex manufacturer for advice as to what options are most common and readily available.
Separation Distance Between Flex Circuit Cores
When a product requires two or more cores, there is a need to define in the specification what the spacing requirements are between cores. The spacing can impact product performance (physical and electrical) and, most obviously, thickness. In some designs, the spacing between flex circuit cores may be filled with dielectric material, but with other designs the dielectric between flex cores in the flex area may be omitted to assure maximum flexibility (Figure 2).
Figure 2: Bonded vs. unbonded flex.
If the core layers must be unbonded, this should be noted in the documentation. Those areas where bonding is to be avoided should be identified in the design artwork package. The unbonded areas must have a coverlayer applied to each exposed side (Figures 2 and 3). In laminated areas, it is not required and arguably a liability when plated through-hole reliability through the assembly process is considered. Obviously, in areas where interconnection is required between multiple layers of internal circuits, a dielectric is required as shown in Figure 2. In the next installment we will continue this three-part series by addressing circuit layup symmetry, designing for bending, and controlled impedance.
Dave Lackey is vice president of business development at American Standard Circuits.
Anaya Vardya is CEO of American Standard Circuits.
This article originally appeared in the April 2018 issue of Flex007 Magazine, click here.