Issue link: https://iconnect007.uberflip.com/i/972396
36 FLEX007 MAGAZINE I APRIL 2018 that it's not just connect the dots, but there's mechanical things that they need to under- stand more about bend radius and stuff like that but they just haven't had a lot of experience. Finstad: I would like to sec- ond that. All the differences between rigid and flex are important. It's just there are so many designers that have only done rigid and they just come in and think, "Well, I'll just use the same rules for flex." And you must coach them that there are a lot of differences. There's more differences than similarities between rigid and flex cir - cuits. McCurdy: We recently did a project for a head- set that the engineers at the company had come up with and wanted to do in rigid-flex, and by the time we got through the project it ended up being so expensive. There are trad - eoffs there with trying to connect two things together that maybe a piece of flex between two rigid boards in a consumer device would have been a better solution. And understand - ing those tradeoffs would have saved them a lot of money. They went down the wrong trail. Miller: Yeah, I would say, too, that IoT has had a big influence on the use of flex and espe- cially from the application of direct attach to flex for components, again it's the DFM issues, but the designers just are not thinking about it in terms of robustness and the application. They're using flex typically for a reason and that flex puts stresses on solid joints that aren't normally there. So again, it's the meld - ing of electrical design and mechanical design on steroids at the PCB and flex design level. Matties: From a designer's point of view, are there component issues that they need to be aware of when dealing with flex? Miller: Well placement, location, proximity to the bends, those are things that you need to be aware of as they're going to be a stress put on that area. Shaughnessy: It sounds like what all of you are saying is that there are a lot of ques- tions about flex. There's a lot of uncertainty at each level of the process with flex. Finstad: Yes, absolutely. Even the SMT stuff. I was in Boston meeting with a customer last week and he asked, "How close can I put the component to the edge of the board? And how close can I put it to the flex rigid transition area?" I gave him the numbers that I felt comfortable with. He wanted even closer. You're always getting pushed. They're trying to use up every square millimeter of space and they don't want to leave anything open. Shaughnessy: In what markets do your flex typically wind up? What segments of the industry? Finstad: The first 27 years it was all military, avionics, aerospace, implantable medical, and now it's high-end commercial, cameras, cell phones and headphones are full of flex now. It's going everywhere; GPS, anything that's handheld is going flex. It's probably easier to tell you what doesn't have flex in it than what does. As I said, when I worked at Minco, everything we did was Class 3. I don't know if I ever saw a drawing in 27 years that said Class 2 on it. If it did, you still got Class 3. It was more difficult and expensive to try to seg - regate a single run and run it to a Class 2 than just run it at Class 3. FCT never competes against Minco because we do Class 2 virtually across the board. Once in a while we'll do a little bit of Class 3, but rarely. The long and short is, anybody who can build in Asia is going to build in Asia, and Scott Miller