Issue link: https://iconnect007.uberflip.com/i/1512857
14 DESIGN007 MAGAZINE I DECEMBER 2023 I do hear what you're saying, and I agree with it. I just don't know if it's valid from a practical standpoint. is leads us back to unconven- tional board outlines. How do we make these odd board outlines manufacturable, espe- cially if we don't know where our designs are headed? How do we incorporate DFM into board elements? Shaughnessy: It sounds like unconventionally shaped boards bring their share of DFM and DFA issues. Moyer: Yes, and design for test (DFT). With a standard rectangular board, you can get all your probes in there, and your flying probe tester is no problem. But how do you test for a one-off board? How do you get a one-off, weird-shaped board into a flying probe tester to do your post-assembly test? You have to say, "Hey, assembly house, don't pull this thing out of the break-off rails until you've completed testing." Dack: SMTA publishes its TMAG specifica- tion on testing, which is widely used and rec- ognized as the definitive explanation. I called SMTA and purchased the TMAG specifica- tion, and I've done a presentation on it. is is what many companies are using to design test fixtures to the correct criteria. With a complex shape or geometry, now you have an in-circuit test problem and you have to deal with it a dif- ferent way. Moyer: at's right. ere is another point I'd like to make. As we move toward 3D-printed board structures, there are huge challenges for the designers because now they have to start thinking like 3D mechanical engineers. It's no longer the same process technologies. It's no longer drill and plate; it's literally composite metal as you're building up. I no longer need a vertical plated through-hole structure. I can have my plated structure from layer one to layer two, just be printed at 9.7 degrees, if that's the angle I want for this weird 3D structure. Dack: I agree. Design is where it all starts. Kris, we've been talking about CAD data, and generally speaking, CAD data is nominal data. One of my pet peeves is that board design- ers need to consider tolerancing. What do you think about that? What's your take on the importance of understanding tolerancing on these processes? Moyer: Understanding those tolerances is essential, especially with these oddball shapes, because a lot of these oddball shapes are going into consumer goods such as gloves or things like that. It's not just fabrication tolerances that must be considered, but environmental toler- ance as well. You must understand both your expectation as a designer and your fabricator's capabilities. Talk to your fabricator early and oen, as well as your assembler, to make sure you understand their properties. Here's an example I use in my military design class. You have a board that must fit into an extruded aluminum housing and it's designed for a board that is 39 mm wide by 1.6 mm thick. But if you make that board exactly 39 millime- ters wide, and then you expose it to extended operating temperature ranges like the military sees, the board will expand faster than the alu- minum, get wedged in, warped, and mucked up in there. Designers need to understand not just the fabricator's capabilities, but what your tolerances are, then consider the environment where you expect your board to be. Is it some- thing that's kept in your home in a controlled environment of 50-80°F? Or will it be for a military product placed in -55°–+105°F? If it's going into outer space, what are the require- ments, and what effects do those environ- ments have on the geometries and tolerances and physics of your board? Shaughnessy: Good stuff. Thanks, Kris and Kelly, for this conversation. Dack: anks, Andy. DESIGN007