Issue link: https://iconnect007.uberflip.com/i/1389320
16 SMT007 MAGAZINE I JULY 2021 panies. is is because they're using a ROSE tester metric which doesn't detect small pock- ets of contamination, that doesn't determine which components are the hardest to reflow due to thermal mass, or whatever it might be, and then being unable to fine-tune the process to that one failed area. I think CMs still struggle transferring the qualification package over to the actual prod- uct, just because it's so different. Johnson: And having the skill set to do that? Camden: Yes. And there's still the issue with brain drain in the work force. We still see man- agers cutting the floor staff who have been building boards for a long time. Take that tribal knowledge away and it really becomes lost— until somebody else learns what the prob- lem is and how to fix it. We see so many peo- ple move from one company to another, taking that knowledge with them and leaving a void behind. Johnson: Cleanliness issues are going to affect reliability and build quality in numerous ways. Could you list out a couple for us? Camden: Yes. It's one of our specialties. When you're building an assembly there are so many opportunities to add contamination; you take these 100 different touchpoints, and if you don't know how to look for each one of them, you can have a failure and never know where it came from. Cleanliness is reliability in terms of operation in the field, and it's just so impor- tant. e first thing you're going to see is electri- cal leakage; probably half of what we do is "no trouble found" conditions. It's the Florida vs. Arizona conundrum, where 10 of them were sent to Arizona and never had a problem, but 10 of them were sent to Florida and every one of them failed. e availability of the atmo- spheric moisture in your end-use environment is such a critical parameter when you're man- ufacturing your board so you know how clean your board needs to be and what levels of resi- dues are allowable. Sometimes condensation on your board will cause it to fail. You say, "We've got a failure here. Let's send it back to the manufacturer." e manufacturer puts it on their desk, and it doesn't fail. Now, they've got the customer swearing it failed and the CM says, "No way it failed." What we find out is that the board is on that hairy edge of failure. With electrical leak- age, you're really lucky if you get a constant failure. at means there's something we can do some testing on to determine exactly what's causing that leakage. ese issues start with electrical leakage, and then if it's allowed to stay in service with this on-again/off-again electrical current, you grow dendrites. Now you've got a hard fail- ure. at's really the second thing, and we're still seeing issues with these electrical leak- age paths happening under conformal coat- ing. e boards aren't being prepped in a way that they're ready to take a conformal coat so are still able to absorb moisture. ere are so many ways to contaminate the board, but the end result is usually some sort of electrical leakage, unwanted voltage; you could even get EOS from that. ere are a lot of different ways to make the board fail with contamination, but it just ends up with electrical issues. Johnson: Eric, is this a shortcoming of the cur- rent cleaning technology? Camden: e technology is fine. It's just a mat- ter of each individual CM—and the opera- tors within that on that line—knowing how to set up a good wash process in the first place. Understanding how to set up the equipment, how much temperature it needs, belt speeds, and how to optimize a wash process—and then verifying that using different methods of test- ing, including just standard burn-in type testing where you put your live product in an environ- mental chamber and see how well you clean.