DuPont, Taconic and PFC Team Up For High-Speed Flex
At DesignCon, I sat down with three flex circuit specialists: Glenn Oliver of DuPont, Tom McCarthy of Taconic, and Steve Kelly of PFC Flexible Circuits. Our discussion covered a lot of territory, most notably the findings they described in the paper they were about to present later that day at DesignCon, and the future of flex, as they see it.
Barry Matties: Please start by telling me a little bit about your paper.
Glenn Oliver: One thing that is not well known is that flexible circuit technology can be applied to high-speed and high-frequency data transmission. DuPont, as well as Taconic and PFC, collaborated in this effort, because there's really no one solution provider who can provide the entire spectrum of what's needed.
For instance, one of the products that we make at DuPont is Pyralux® copper clad laminate. This product has a very good, long-standing reputation in the flex circuit world. What we found, though, is that we didn't really have a low-loss adhesive that was really tailored for high-frequency applications. Taconic actually has that piece of the puzzle and I’ll let Tom talk about that.
The other thing is that fabricators are an incredibly important piece to it, and one of the fabricators leading the way is Steve Kelly at PFC, who really has a lot of understanding and expertise in the area of high frequency flex. This is actually low-frequency for Steve, because he works a lot in the optical business. It's really in his wheelhouse and so it's a natural collaboration, but I'm going to defer to Tom to tell us a little bit about the adhesive technology and how that enables what we're doing.
Tom McCarthy: Taconic specializes in RF and microwave frequencies. We sell products out to 77 GHz, so most of our core competency historically has been rigid boards or multilayers or microstrip, double-sided boards. In 2008, we developed a product called FastRise™, which is a non-reinforced pre-preg. This was developed for the rigid marketplace, but it looks a lot like DuPont's offerings in the flex market space, in the sense that it's based on film technology. At some point we decided to try to make a flexible version of our FastRise™ prepreg that could be used in the flex industry. Again, it looks a lot like the approaches that DuPont takes with thermoplastic films in the flexible marketplace.
The frequencies have been going up incredibly in the rigid marketplace. The frequency has been going from 28 Gbps to 56 Gbps in the digital world, but in the automotive world, collision avoidance radar is 77 GHz, so we're used to tackling these kinds of frequencies. Now it's just been more taking a closer look at what the flex world needs, because the same demands for higher speeds are taking place in the flex world. So it puts a lot of pressure on all the material vendors and the fabricators to create structures that can meet higher frequencies.
Matties: Steve, as a fabricator, what are the challenges that you're faced with?
Steve Kelly: Primarily, we've been working in the high-speed market since around 2000 and at that time we started working with various companies to build 10 GHz circuits. That was pretty standard technology back then. DuPont had the materials, other people had the materials, but DuPont had the best stuff. We made that for quite a while and then in the last two to three years people started to move into primarily 18 GHz up to 50 GHz. These are test equipment manufacturers and what I call the optical companies—people who are making the network systems for 4G−5G and so on. We started to look into these new materials. DuPont came up with a material called TK three or four years ago. We started to build with that. We are happy with that, but then eventually you go from “Now I want a multilayer” to “Then I want a rigid-flex,” and then you've got to start bonding it all together. We needed the Taconic bondplys and DuPont TK, or AP material and some kind of cover-lay to put it all together. We've been working back and forth now for, what, two years, Tom?
McCarthy: Yes, a couple of years.
Kelly: He's made a couple of different products, and the new stuff seems to be working pretty well. It's not the easiest stuff in the world to work with, but as life goes on you get better and better at it. It's not normal fabrication.
Oliver: The ultimate would be encapsulating your circuit in Teflon®, like in a high frequency cable. But the things that make Teflon® great from a signal integrity standpoint also make it really challenging from a processing standpoint. A huge advantage of Teflon® is that it is inert to most chemistry. That’s great, but how do you remove Teflon® from where you don't want it after lamination? The real enabling part and why this collaboration is so noteworthy is that DuPont obviously has the expertise with Kapton® and Teflon®, but Taconic has a missing piece of the puzzle in that they have a really good low-loss adhesive. This adhesive doesn't require the high temperature processing that's required for fusion-bonded Teflon® fabrication. It's not the “magic bullet” for every circuit design, but it’s a great solution when you need high-frequency and you need the flexibility for a printed circuit.
Matties: Where is the limit before it just becomes an optical solution?
McCarthy: We saw a presentation today on 56 Gbps, and they're not using the lowest-loss materials or even the flatter copper that is coming into the marketplace. And they're not using the newest treatments of oxide treatments on the innerlayer copper. Innerlayer copper usually gets some type of oxide or alternative oxide, so there's new technology in treating the innerlayers. I think you're going to see copper continue to keep going out further and further.
Matties: So there is head room?
McCarthy: Yeah, absolutely. There's more room to grow.
Kelly: You know, the military has been using blind and buried vias forever. In the telecom world, they try to get away from blind and buried vias and a lot of microvias, because of the cost. It's expensive to sublam, but the military has been doing three, four, five sublaminations forever. It's just a matter of time before all of this stuff trickles into telecom.
Matties: From a value point of view for your customers, is this expanding your market share or giving you a competitive advantage?
Kelly: It certainly gives us a competitive advantage because there are not very many people who can build this. We're probably at the forefront right now. People will catch up as this becomes more available, but PFC has always tried to stay a little bit ahead of the curve in terms of densities and new materials. Quite frankly, for a long time, there weren't a lot of new flex materials on the market. Now there's a huge proliferation and we're playing with quite a few of them. We're onto our fifth try of a new high-speed cover-lay this year already. The first two failed miserably, the third one looks good, and then I've got a new one I'm trying in the next couple of weeks. This is all kind of coming together, because our customer base needs this stuff yesterday. We're already shipping product, not a ton, but we are shipping, and it's working in high-speed applications. They're running the stuff at 28, 40 and 50 GHz, and they're happy. It's not only the speed, but they love the low loss and the low moisture absorption. That's the big thing. It makes a lot of difference to them.
Matties: Just for a little context, why don't you just give us a little overview of PFC.
Kelly: I started the company in 1997. Like any flex startup, we struggled for a while at first. In 1999 I was called down to Waterloo [Ontario, Canada] and I met with a company called Research in Motion. They were 35 people at the time, and for the next six years we built every flex circuit that they wanted.
So, we basically grew with them. In 2001, they came to us and said, "We hate assembling these things, they drive us crazy." They didn't fit the same profile as the rigid boards, so they were doing all their assembly in Waterloo. We got into the assembly side of the flex circuit business and made good money from them. They were a great customer. We primarily poured that money into high-density, next generation stuff, because in our opinion, that's where the market is going.
Research in Motion eventually went offshore, but for three years, from 2005–2008, their phones didn't even use flex, which was kind of a surprise, so we didn't care anyway. Then they got back into it. You look at any phone nowadays and they're just filled with flex circuits. Apple just dominates, in some cases, the flex circuit market. We continue to push that high-speed stuff and it’s still important to us. We do a lot of medical. The company is very successful.
Matties: Congratulations. And Tom, why don't you give us a little overview of Taconic?
McCarthy: Taconic was started by a person who worked at DuPont, left DuPont, and later built a factory on farmland that his wife owned in the middle of nowhere, east of Albany, New York. In the ‘60s it started out with fiberglass-reinforced Teflon® composites. In the ‘80s, it was fiberglass reinforced Teflon® composites for the circuit boards. We've been building Teflon®-based circuit materials for 35 years now. We want the performance of Teflon®, but we don't want you to even realize there's Teflon® in it. We want to make the Teflon® piece invisible, aside from the electrical parts. That's my primary job—to make it fabricate like it's not Teflon®.
Matties: And Glenn, we all know about DuPont, but why don't you just tell us a little bit about what you're doing over there?
Oliver: I do a lot of things at DuPont, and I'm not going to go through the whole history of DuPont, but I can say that a lot of people see DuPont as a chemical company. I think it's safe to say that we're definitely beyond that now. In fact, really it's more of a science and engineered materials company now than it ever was. I can definitely say that we're a company that is constantly changing. We've evolved significantly over the past two centuries. It started in 1802, so it has a very long history.
It has a good brand name and it should, because it's a good company to work for. I can say from my role there, what I focus on is this area of high-speed and high-frequency flex. It's still really a brave new world, because the flex circuit market is very small compared to the rigid circuit market. The knowledge base in terms of what designers have and what they understand to be the options is even less represented.
I guess you could say that one of the things I should have on my business card is evangelist for using flex in high frequency. I have a lot of passion for that business and I think it's an area that's going to win in the market, because there's a lot of untapped potential, as Steve and Tom pointed out.
Our paper this afternoon is going to show a lot of misconceptions that people have about flex circuits and hopefully we may be able to dispel them. That's one of the exciting things about my job because often designers have no idea that flex circuits can be a solution to their problems.
Matties: It sounds exciting, and you also have good timing. What do you guys think about the market for this year?
Kelly: The flex circuit side is pretty strong. We are a North American fabricator, and we’ve almost doubled in the last five years.
Matties: That's great. How many employees do you have?
Kelly: We have 110. We’ve just seen more and more business. Eventually, the high-volume stuff will go offshore, it always does, but the specialty stuff, the high-volume, low-mix, it's always going to be here. The new stuff is going to be here and then it migrates. Packaging is everything. They need everything smaller and smaller, lighter and lighter. If you look inside the Apple watch, there are 10 flex circuits. There are no rigid boards.
Matties: There's not even room for 10 circuits in there!
Oliver: We want them big!
Matties: So much easier to deal with our aging eyesight!
Kelly: It costs me more to count them than it does to build one. The iPhone6 has, I think, six or seven, for example. And 10 GHz is becoming very over the hill and now everybody is moving into 18, 28, 100 GHz. We've even had phone calls asking, "How do I build a 100 GHz flex?" I say, “Well, I might be able to build it but how are you going to test it?”
Oliver: There's a lot of flex in test equipment. That's a very important market for Steve. It's a testament that you can use high-frequency in flex because a lot of test equipment uses flex in the probes and then in various other aspects of testing.
Kelly: This stuff starts up there at the test guys and eventually migrates down, just like anything else in this world. Then it becomes mainstream and this is on the cusp of that, and it will start developing.
Matties: This is great. It's nice to see that the edge is being pushed out further and further. Good for you guys. Congratulations on your paper, too.
Oliver: Thank you for the interview. We appreciate it.
