Issue link: https://iconnect007.uberflip.com/i/1233021
14 DESIGN007 MAGAZINE I APRIL 2020 tic impedance of the traces need to be. I create an appropriate board stackup to accommodate these signals and select vias accordingly. Also, what type of finish do I need on the board? What are the copper thicknesses? Can I use a standard laminate, or do I have to use some hybrid lami- nate or an embedded capacitive substrate? What mechanical items do I need to address, such as specific connector locations, tooling/mounting holes, board outlines, and any special require- ments? Do I use the components I have in my library or make new ones? I usually spend a lot of time working on my library of components. Nolan Johnson: You mentioned you do your own parts in your libraries rather than using any outside services. Young: Actually, I have a librarian who is re- sponsible for creating the parts, and I'm re- sponsible for reviewing them and making sure they're correct. Johnson: On average, how much of your over- all design hours are devoted to library manage- ment? Young: It depends. I also have my own com- pany and am in the process of integrating with The Goebel Company. I've been working on my current master library for about 12 years. Early on, I spent about $300,000 on setting up my library. Now, approximately 20% of my budget is dedicated to library development. I try to do it upfront. I'm always looking at vendors' websites to see what they have for new components and packaging technology. I have my librarian on an annual retainer, which gives me on-demand library services. Part of my business model is that I'm asked to do things very quickly. I might do a board that's 5" x 7" and have 100 parts. I can usu- ally do something like that in under 30 hours. Sometimes, I'm lucky, and all the components I need are in my library. Shaughnessy: That's a good person to have on retainer. Most designers I know hate to spend too much time in the library. which has shaped how I approach the issue of costs and profitability. It also depends on the environment that you're in. I always ask myself basic questions, such as, "What are we trying to do? When do we need to do it? How are we going to do it? Which market are we addressing? What do we think our budget is?" I say "think" because the budget is usual- ly guesswork at the beginning. I think about what it would cost to do something based on metrics and previous design experience. Typi- cally, I take into account planning, design, test, manufacturing, and sustaining costs. I add up the time and materials and then push it into some type of overall estimated budget. As the project/program progresses, the budget is ad- justed and/or addressed accordingly. Shaughnessy: Can you tell us how you ap- proach it from a PCB design perspective as well as an overall project perspective? Young: Sure. When I'm thinking of board de- sign, the things that initially come to my mind are as follows. What are we trying to do? Is it a test fixture board? Is it a product or an in- strument of some sort? Is it something else that goes into a box? Is it a computer? Break- ing it down further, what type of interfaces do I have? Do I need analog, digital, power, or RF I/O? What are the number of interfaces on the board? What type of power rails are available from the outside world? What type of connec- tors do I need to be concerned about? As I dig deeper, I consider the type of signals I have to work with and what the characteris- In my career, I have developed a systems view of what's happening, which has shaped how I approach the issue of costs and profitability.