Issue link: https://iconnect007.uberflip.com/i/1535183
MAY 2025 I DESIGN007 MAGAZINE 19 Another big challenge is understanding the steps you need to document. For example, in a fabrication drawing, I've seen many design- ers use incorrect standards in their statements. ere are several material standards, such as IPC-4101, IPC-4781, IPC-CC-830, and IPC- SM-840, but none of them tell the vendor how to apply that material to the board. ey sim- ply tell you the properties, the viscosity, the electrical resistance, etc., of these materials. I've seen too many designers tell a fabrica- tor to apply a solder mask in accordance with SM-840. at's not the right way to document it. ere is a learning curve for how to create your documentation properly. How much information do fabricators want? Fabricators say, "Don't tell me what to do; just tell me what the requirement is. What's your pass/fail criteria? I will get you past the line if you just give me the criteria, but if you tell me step by step what to do, I will do exactly what it says because that's part of the IPC require- ments." e master drawing is the controlling document, so if it says to do it that way in the master drawing, that's what the fabricator will do. Now you have basically superseded the fab- ricator, assembler, or tester's expertise. I've heard CAM engineers say, "Don't tell us FR-4. Just say traditional or standard laminate." Correct. I actually go right to the IPC-4101 material standard, because there is now such a wide variety of laminate formulations that could be considered standard at Fab Shop A and not standard at Fab Shop B. You could have two vastly differing performances of your board. I won't say FR-4; instead, I'll say, "I need a material with a multifunctional epoxy- woven E-glass reinforcement. I need this dielectric constant, and this dissipation factor." I'll give you the properties that I need. You can pick manufacturer A, B, or C for the laminate materials, and as long as you meet those prop- erties, I know my board will function to my design. at's a big challenge in creating doc- umentation: designers using colloquialisms as opposed to technical definitions of what they're actually trying to achieve. It's very dif- ficult to overcome. Is there anything you'd like to add, Kris? First, make sure that the level of documenta- tion you are creating is sufficient for the prod- uct and the market segment, because you can cause problems by being overly defined. If you don't need Class 3 documentation and all these dimensions, why pay for a QA inspector to spend three weeks measuring 50 different measurement points on your board? But if it's a military satellite, sure. Next, understand the appropriate level of information necessar y for your product and your market segment. Understand the lan- guage and how to write all of this clearly. Understand the materials and the data styles when following the IPC standards. Under- stand the class, the documentation grade, and completeness mode, as well as how that information should be conveyed. Give appro- priate tolerances. One example I'll give you of improper documentation is minimum trace width. Many a designer has said, "e minimum trace width I designed in my board, maybe with one-ounce copper, is a 0.005- inch trace." at's great, but if I put that on my drawing and don't list that as a reference dimension, that has all of a sudden become an absolute minimum. If my fabricator etches that trace and it comes in at 4.9995 mils, it's undersized and it's scrap, because I haven't allowed for the tolerance. Designers need an understanding of how to write the notes to take into account the IPC classifications, such as the IPC-6010 series of fabrication performance and quali- fication, to make sure you're not accidentally overconstraining the fabricator or assembler to decrease your yield. Thanks for speaking with me, Kris. ank you, Andy. DESIGN007