Issue link: https://iconnect007.uberflip.com/i/1484263
NOVEMBER 2022 I PCB007 MAGAZINE 33 Johnson: e dimensions you're trying to man- ufacture are now inside the tolerances you used to allow. It changes what you can do with your equipment. Patel: Exactly. Some fabs have 20-year-old equipment; plating tanks never get improved and it's very traditional. You can't just use the same equipment you used in the 1980s or '90s to produce the same boards that you're doing now. You must innovate that type of fluid dynamics or purchase the right pieces of equip- ment to keep up. Johnson: Based on your experience, is there room to optimize existing equipment and dial it in so that you can do HDMI? Are you looking at a new line to get those UHDI dimensions? Patel: I think that becomes a production issue. Yes, you can finesse your equipment to do pro- totype work, low volume, four or five panels, with someone taking care of the system while it's running. But can you do 100 panels of this in that same format? No, I don't think so. Johnson: How are you transitioning the manu- facturing floor? I'm sure it's a step-by-step pro- cess. Patel: Right. Yogen Patel has always looked to simplify, to eliminate extra steps with tech- nology that improves it. One example is liq- uid photoresist instead of dry film. at's a big thing because our coated resist is very thin. When you go to UHDI with dry film, you can get away with a 0.5-mil dry film or even a 0.3- mil for imaging. Of course, it gets very hard to handle, but people are using this already. at was our original roadmap. Direct imag- ing becomes very important, and the type of direct imagery you use, too, because some of these LEDs have a limitation on the pixel size. It becomes very important which machine you pick for that type of work. For us last year, we picked up a lot of pieces that will help us in this type of work with the direct imager, the Indu- bond machine, or very consistent pressing, and then the DIS bonder. You don't have to pick that exact equipment. But the idea is that you're trying to improve the layer registration, trying to reduce the use of thick film plots so that you have direct pho- tolithography for sharper imaging. You need even pressing so that you don't have variation between the manufacturing panel thickness. All those things are very important. e next tool you need is a very solid laser drill that reduces the amount of carbonation that you get because that's a very big quality headache. Doing UHDI without a laser drill is just very difficult when you're dealing with sequential drilling and blind vias. It's possible, but it's not easy. I really like the idea of the green pico laser because we're seeing evaporation vs. carbon- ation of material. I know TTM has a verti- cal plating line, but we've been working on a vertical plating line for five, six years now. It's almost ready, but that's the step that we're going toward where it's knife edge inner anode system for extremely even plating. We're doing things to avoid extra plating on the sur- face. We're looking at more automation so that there are fewer chances of any material mechanical issues. Much of it is soware as well. We invested in pre-CAM soware for Doing UHDI without a laser drill is just very difficult when you're dealing with sequential drilling and blind vias. It's possible, but it's not easy.