Issue link: https://iconnect007.uberflip.com/i/1000349
24 SMT007 MAGAZINE I JULY 2018 Keeping: I completely agree with that, and remember, usually the designer will have, let's say, an electrical background or design knowl- edge. They're not assembly experts, they're not a ruggedization expert, so they've got one scope or one set of glasses that they're looking at to solve the problem and, of course, they're going to use that to their full knowledge. But if they don't share that with that other expert, then they've only got half the information they need to get the best design for the first time. Willis: I think that every company should have some form of DFM review, even if it's just 10 people. You look at the product, you look at it as it goes through manufacture, and you make a simple note about what worked and what didn't. If you pass it to another department for cleaning or for coating at the initial prototype stage and get them to sign it off in some way to confirm it is manufacturable or to confirm what needs to change.. That sort of stuff doesn't exist in a lot of companies, going back to find out who looked at this, who decided on that. That information's not available, but when you're creat - ing a new design, however you're going to do it, all the infor - mation needs to be kept together. And so often when you do failure analysis, as I do all the time, you say, "Okay, where is that product going to go? What procedures need to be followed? Where is that infor - mation? Where are the test results?" Just recently, I had a corro- sion failure. This product was designed to be in an automotive application, exposed to mois- ture, and they didn't have conformal coat- ing included even though it was obvious that condensation would take place. It's not obvi- ous it's going to have a failure, but it's obvi- ous the condensation was going to take place. Because the customer did not ask for coating, no one questioned the potential for failure. Potentially, we might have corrosion, which they've actually got. I said, "Okay, what did you do? Did you test it?" "Yeah, we tested it, this is a special occasion our customer wanted us to test it to." But it's not relevant for what they're doing. That company is taking reliabil- ity of that product they're building without fully understanding, or the customer under- standing the potential risks. Now, there's some argument whose fault it is. I would say, "Well, sorry guys, you haven't followed it through properly and that's the reason you've got a problem." Las Marias: Speaking of particular processes in the assembly—material handling, paste print- ing, component placement, soldering and conformal coating—Jason, what are some of the best practices that you can talk about? Keeping: I think you had a very good list there. If I look at your first topics here, material handling, paste printing, component place- ment, reflow, solder- ing, they're not my areas of expertise. I'll probably leave that for Bob to talk about, but all of these processes do have a common aspect that does lead into my area of expertise, and this is assembly board ionic levels. With the conver- sion to no-clean, I just want to say right there that's a myth that was started. The assembly cleaning was no longer required, however we still have low levels of ionics present. I prefer to say "low solid materials". Based on the format of these processes these can even still be very high and even exces- sive to the point of source ionic failures later in the process or in the field, such as mate-