Reading time ( words)
We asked for you to send in your questions for Averatek’s Tara Dunn, and you took us up on it! We know you all enjoyed reading these questions and answers, so we’ve compiled all of them into one article for easy reference.
Reshoring of High-Volume IC Substrate Production
Q: Do you think IC substrates (BGA, CSP, flip chip, etc.) will ever be produced in high volumes in the U.S.? The Asian capacities are currently full and expanding.
A: That’s a great question. First of all, I think the concept of high volume is relative, and the definition may be something different for everyone. There is quite a bit of effort ongoing at this time to bring this technology back to North America. These endeavors are exciting for the industry and important to the Department of Defense and its mission of providing security for the U.S. These “reshoring” efforts have begun with limited capacity in comparison to the overall volume in the market, but as development progresses, we might find that North American manufacturing may be more competitive than we realize, thus drawing additional production back to North America.
The Thickest Flex Layer Available
Q: What is the maximum thickness of a single flex layer, not just for a test vehicle but a flexible circuit layer manufactured commercially?
A: There are a lot of different ways to answer this question. There are examples of flexible circuits being manufactured with 10-ounce copper bonded to dielectric. In this case, the “body” of the flexible circuit is typically etched to a thinner copper for flexibility and leaves the contact fingers with the thicker copper.
Looking at what laminate thickness are commercially available for fabricators, Dupont provides a laminate with 0.005” dielectric, bonded on each side to four-ounce copper, with 0.003” adhesive. That is an extremely thick laminate, at 0.0222”. The most common flexible laminates are much thinner, with half-ounce or one-ounce copper bonded to 0.001” or 0.002” polyimide, resulting in thicknesses of 0.0024–0.0068”, depending on adhesive requirements, dielectric thicknesses, and copper weight.
Fabricator Feedback to PCB Designers
Q: Why don’t flex and rigid fabricators provide more feedback to designers, especially if it’s not good design and engineering work?
A: This is an interesting question. My initial tongue-in-cheek reaction was that fabricators are not often asked to provide feedback. In my experience, fabricators are very willing to provide feedback on PCB design, flex or rigid, and that conversation is something that I would always encourage. These conversations are always more impactful when fabricators are involved early in the design process.
Discussions about stackup and design for manufacturability concerns—especially for designs pushing the limits on multiple different fabrication parameters that require trade-offs—or for designs with concerns about flexibility and robustness of a design are critical and can have a significant impact on cost and reliability.
Unfortunately, with the rapid pace of the industry, those conversations often happen only after a design is complete and sent to the fabricator for manufacturing. At that point in the process, time deadlines often loom, and it can be time-consuming and expensive to rethink that design to improve manufacturability.
I would challenge both designers and fabricators to take a step back and find ways to better communicate about DFM issues early in the process. There is a lot of great content available, but that cannot replace conversation, or even a facility tour, as a learning tool. Reach out to your fabricators early in the design and ask their opinion.
Flex and Rigid-Flex DFM Advice
Q: Flex designers don’t seem to pay much attention to a fabricator’s limitations. What would you advise new flex designers as far as DFM?
A: Knowing your fabricator’s limitations is important, whether that be for rigid PCBs or flexible circuits. And this is especially important with flexible circuit designs. There is a wide range of capabilities across the supply base. Some focus on simple flex, single-sided, double-sided, or low layer counts. Others focus on high-complexity technology, including complex rigid-flex and bookbinder technology. It is critical to really understand your fabricator’s capabilities and design to those capabilities.
As important, in a general sense, is to understand that flexible circuit fabrication, while similar to rigid PCB fabrication, has subtle differences that impact things, such as minimum line width and space and hole size. Not only are the layers thinner, with considerations needed for handling and processing, but the materials themselves are less stable than traditional rigid materials, with subtle shifts in material size through processing and changing environmental conditions. These things impact registration during fabrication, and feature sizes that may be common in rigid PCB fabrication may not be the “standard” for flexible circuit designs. Communication is king.
The Future of LCP as a Flex Substrate
Q: Will the surge in liquid crystal polymer use for cellphone antennas “trickle down” and allow LCP to be used more often as a conventional flex substrate?
A: That is a great question. I am certainly seeing an increase in conversations about LCP materials and their benefits not only in cellphone antenna applications but also in medical applications requiring biocompatibility. As with anything cellphone related, there will be a handful of fabricators that use the materials in high volume and over time that trickles through the supply chain. In the case of LCP materials, the fact that there are specialized fabrication parameters that manufacturers need to employ when processing LCP and that the conventional flex substrate market is robust leads me to believe that while we will see an increase in demand for LCP, the material likely won’t displace polyimide as the mainstream flex substrate.
Maximum Number of Flex Layers
Q: What is the outside maximum number of flex layers that have been used in an actual product?
A: I could answer with a number based on just my personal experience and memory, but I think it is important to include some context along with that. The key issue with the number of layers in a flexible circuit is the ability for that flex to function in a reliable manner, over time, as intended, whether that be flex-to-install or in a dynamically flexing application.
“The flex that didn’t flex” is something we joke about, but it happens more often than one would think. I have seen three-layer flex circuits that were not flexible enough to be bend around the corner of the package, and I have seen 12-to-14-layer flex circuits that functioned beautifully in high-flexing environments. Key considerations to improve flexibility such as material selection, using unbonded layer construction and cross-hatch copper patterns vs. solid copper can have a significant impact on performance.
Involve your fabricator early in the design cycle; they have a lot of knowledge to draw from on how to best meet your overall flexing requirements.
Remember, if you’d like to hear more from Tara, be sure to view her Flex007 column series “Flex Talk.”