Issue link: https://iconnect007.uberflip.com/i/1007258
12 FLEX007 MAGAZINE I JULY 2018 life cycles and various envi- ronments. We do a lot of design for reliability work and modeling. Then we do actual testing. What's interesting is that in the last three years, miniatur - ization has really forced us into the flex world, which is basically new to us. We've always had the luxury of relatively big boxes and lots of space to accommodate the electronics, but now our products are portable or wearable. It's all about min - iaturization, so trying to origami your design into a three-dimensional space has been really interesting and challenging for us. We're new to this, and we're learning the design rules and the limitations of the supply chain—where we can push the boundaries and where not to—so it's really exciting work. Shaughnessy: So you were basically forced into flex? MacFadden: Yes, and we've followed a lot of the industry in this vein. I think a lot of people are just really ramping up with flex, at least in the consumer electronics space, for the same reasons we are. But we've managed to avoid it until this point because of our view that flex was more expensive than just getting a board-to-board connector. What we've learned is that's not always the case. Sometimes it's actually cheaper to make a custom flex circuit that can be integrated into a product in a more elegant way. Not only can you reduce space and even mass, but you can save costs. You don't have to buy board-to-board connectors in all cases. Shaughnessy: It sounds like you're having to learn the design rules, which are entirely dif- ferent than rigid. MacFadden: It is, and you know what's really challenging? We get input from our supply base of what their capabili- ties are, and that's really the baseline for our inter- nal design rules at Bose; but every supplier is unique. They have different capa- bilities. They have different material sets, and different sweet spots, so a design for supplier A may or may not be optimal for supplier B, and because we always try to dual-source just to mini- mize risk, we really have to optimize that design so that it's robust enough for both suppliers. Sometimes that means making trade-offs. Maybe we can't make something as small as we want to, or we can't put it as close to the edge as we wanted. Shaughnessy: In the survey for our inaugural issue, you mentioned that it's important to work with the fabricator and assembly people early on. Do you always know which fabrica- tor you're going to use for every design? MacFadden: Yes, so we've gotten to know our supply base really well, and we know what their capabilities are now after having a lit- tle bit of experience. I generally know what that form factor is going to be and what the technology's going to be on that flex circuit. I know which two or three suppliers I'm going to start engaging, so we'll just bring them into the room or have a call with them even before there's been any pen to paper, when it's just a rough concept. "We need this sort of layout. It's got to fold up into such and such form factor." Then, we start asking questions. "Is it better to use a stiffener? Is it better to use double-sided? Is it possible to do double-sided surface mount?" We've had that question come up a lot, and it's really key to get those answers at that point of discussion rather than trying to manipulate a design later. As you know, every design itera- tion becomes more expensive and difficult to change later. Todd MacFadden