Design007 Magazine


Issue link:

Contents of this Issue


Page 29 of 81

30 DESIGN007 MAGAZINE I MAY 2023 It's the same with assembly. We've got the yellow book—IPC-A-610—which covers the acceptability requirements for all the assem- blies. It's full of color pages of the target condi- tions (my column name), acceptable criteria, and nonconforming criteria for each class. e project that you're suggesting might be a daunting project to specify, or be an index for materials used in design because now you're getting down into the granular composition of the materials. Every supplier will be fighting against the competition and tradem a r k i n g , s o they will tweak their comp o s ition just slightly so they can get around the rules. Shaughnessy: It sounds like you're saying that the current methods for specifying mate- rials work pretty well, and if designers get too granular with their metrics, they can wind up over- constraining the board and raising the cost when they don't need to. Yes. I can show you. e two big over-constraining areas on bare board PCB fab- rication specification are mate- r ials, and dimensioning and tolerancing. Design vs. manufacturing is so dynamic and can be subjective; sometimes you have to rob Peter to pay Paul. Holden: The problem in materials is that the suppliers selling materials have a reason to make specifications and information available. When we talk about design, we open a can of worms. What market is the material designed for? What is the perfor- mance, reliability, lifecycle, and cost? Those trade-offs would be very difficult to document. The designers who know the most don't have time to sit down and write it out, whereas the suppliers have that incentive because it helps them sell. A designer is not selling anything. If the designer doesn't care about the board, he can use any 130°C Tg material. They need to know where to care. Happy, l like that. PCB designers need to "know where to care." Here's a quick story: I got a circuit board from a customer and it had gone through years of evolution. It was a very simple circuit board, and it had a carbon- wound through-hole resistor on one side. It had a surface mount part on the bottom side, and another surface mount part on the top side. From a thermal excursion standpoint, they had to r un it t h r o u g h o n e pa s s f rom the t o p s i d e , o n e pass from the bottom side, and then it had to get hit with the leads of the resis- tor. For three thermal excursion passes, we typ- ically recommend that we go to the higher-temp mater ials. We also do c o s t - d o w n e f f o r t s ; we'll make recommen- dat ions through the DFM studies that we do. I said, "How much are you paying for that axial lead resistor?" It turned out to be $3 because they're just not as available as the surface mount parts. We recommended switching it to a surface mount part, putting it on top, taking the other surface mount part on the bottom, and putting it on top. We got all three of those part types on the single board surface, one pass through the oven, lowered the material requirement to 130°C Tg, and saved a ton of money on a board that's being made in the millions. Shaughnessy: We're seeing more young designers now. What advice would you give someone who's looking at these different sources of information about specifying and selecting material? The two big over-constraining areas on bare board PCB fabrication specification are materials, and dimensioning and tolerancing.

Articles in this issue

Archives of this issue

view archives of Design007 Magazine - Design007-May2023