SMT007 Magazine

SMT007-Oct2025

Issue link: https://iconnect007.uberflip.com/i/1539960

Contents of this Issue

Navigation

Page 45 of 81

46 SMT007 MAGAZINE I OCTOBER 2025 components. We deal primarily with OEMs and the top OSATS. That's our core business. We have three different types of optical measure- ment techniques. One is our Shadow Moire, which was created and developed within Akrometrix. We're the only manufacturer of this technology. It uses camera light, glass, and mir- rors to reflect the shape of the different things being measured. We measure components or PCB warpage, whether it's at room temperature or under specific thermal conditions. Another technique is our digital fringe projection, which is used for sudden height changes. It can be used for components mounted on the boards or a fully populated board, both at room temperature and under thermal stress, although it's mostly used in thermal applica- tions. It can be localized, and using the fringe tech- nique is mostly application-dependent. Our third technique is called DIC used to mea- sure CTE, meaning in-strain plane measurement. It involves measuring the thermal expansion of the component or thing being measured. It's a two-cam- era setup, which allows us to maintain location and so forth. Some of our competitors only use one cam- era, which we feel does not provide good accuracy. We talk a lot about thermal properties and how technology is pushing the boundaries of what traditional manufacturing can build. What has changed regarding thermal warpage? Are there new requirements? Hubble: It always changes and quickly. The indus- try has faced many challenges over time. Head-in- pillow was a big driver for a long time in the mobile space. One driver that has always existed in elec- tronics is being smaller, thinner, and lighter. That will always be a driver for this industry. How small can you make your features while still keeping them flat enough to have adherence? It's new in that it just keeps getting smaller. It keeps pushing that boundary. In some cases, it's changed how we test warp- age, even against industry standards for package warpage and BGAs. How do we measure the board or component, and from which side do we measure, which can vary from the norm? We've been work- ing with many OEMs and ODMs so that we're all on the same page as they communicate back to their OSATS on how we do the quality control. The other extreme is large packages. There's so much data in the world these days. AI is a data hog. We have these massive packages that we are, as an industry, trying to put down onto very large server boards. Can these big packages use the same rules for package warpage? How much does the printed circuit board side play into that equation? Frankly, the board side of warpage, in terms of local area, has been somewhat ignored. We've always known it to be a variable, a potential problem, but it's becoming a variable that just can't be ignored. That is what we see for these packages and the roadmap for where they're headed. Let's talk about standards. Are they where they need to be? Hubble: What is there is strong and established, but there is a real need on the board side. There will definitely always be that discussion between the manufacturer and the designer about where that number should be. We know that many manufactur- ers will need to go beyond any industry standard. But as an industry, there is a significant lack of stan- dards on the board side, and we're currently work- ing with industry leaders to respond to this. Things change at such a rapid pace now. Road- maps can be tough to pin down. How do you man- age that variable when it comes to creating the standards you operate by? Hubble: Responding to shifts is part of our com- pany culture. We call it being "quick and nimble." We don't look that far ahead in terms of roadmap because we know the industry demand from our customers will change pretty quickly. Maybe the biggest change is with these massive components. The challenge is how we can heat these massive components effectively. An impor- tant variable in our tool is how to get that heat energy into these large assemblies, just as a reflow Neil Hubble

Articles in this issue

Archives of this issue

view archives of SMT007 Magazine - SMT007-Oct2025