Issue link: https://iconnect007.uberflip.com/i/1161956
28 SMT007 MAGAZINE I SEPTEMBER 2019 cal migration notably, which takes so many dif- ferent forms. Matties: And with HDI coming into a stronger play in the market, are we see- ing more failures? Naisbitt: Yes. That was why we became involved with the conformal coating because it helped to pro- tect the circuitry. That was the first introduction I had into what makes electronic circuits reliable. That was the key feature of our business. I became more involved at that point to figure out how to control the production process. One of the main elements of this is that all fluxes leave residues. Surface insulation resistance (SIR) testing was developed and used to try to evaluate whether it was acceptable or not when you were using a no-clean process. With a new process and test, then you need to develop the standards. In 2007, I finally managed to publish the first meaningful documents, which was the IPC test method, and coincidentally, the IEC standard. From there, we have developed even further. The issue came into play around 2010– 2012 as the industry was migrating more into increased packaging density and insulation resistance testing became a necessity if you needed to establish product reliability. But the question that kept coming back to me is how do you control a process? You can't use SIR testing because it takes too long and it's too cumbersome and complex. But what you can do is use that to establish a reliable production process, and then take it from there. You need to be able to measure any mobile ionic species left on the surface that can compromise perfor - mance because if you put electricity through a product in a damp environment—and you have ionics present as well as the perfect electrolytic cell—funny stuff happens. During this time, there was some research into this subject, and a stake was stuck in the ground that said anything above 1.56 micrograms per centimeter squared of sodium chloride equiva - lent is bad and below is good. Now, unfortunately, the industry started with- out any hesitation to adopt the term "cleanli- ness testing," and it isn't. It never was. The challenge is the simplified ion chromato- graph test, aka ROSE test- ing, doesn't differentiate species; it simply looks at a value of conductivity of the solution as to whether or not it falls within acceptable limits. I have spent 25 years of my life trying to explain this to people. You have to stop thinking of it as a cleanliness test because it isn't, but it's an excellent method to control a process. Around 2013, we received a call from Robert Bosch. They require Six Sigma for their produc- tion process. It's automotive electronics that are ultra-high reliability and safety-critical. We took on the project, and in five years of work, we achieved Six Sigma verification uniquely for the first time in the history of this entire sub - ject. To do this, the instrument needs to be suf- ficiently capable of detecting changes in the process as fast as possible. We use a merit-to-fit algorithm. It means that we were able to give an accurate go/no-go answer within typically three to five minutes. What we're looking for are only those mobilizable ionic species. Clean - liness is something entirely different. Matties: After your five years of work, the cur- rent pass-fail approach doesn't make sense as a standard. What was the response, and how do you carry that change in the standard pro- cess to the industry and get support? Naisbitt: There are two answers to that. One, I'm silly. Matties: That was a short answer (laughs). Gen3's CM+ Series, the worlds first combined ROSE and PICT Ionic Contamination Tester.