Issue link: https://iconnect007.uberflip.com/i/1309864
NOVEMBER 2020 I PCB007 MAGAZINE 37 people. We need to get some prices from the OEM. Matties: The fact that you convinced the tool company to integrate it into their tool means they recognize this as a technology that will be uti- lized. That's a good invest- ment on their end. Dickson: For us, that was huge. As the PCB suppli- er, we started building the structures so that people could understand what the structures do, and that took maybe two years. Now, we have Ca- dence, Zuken, Mentor. We're primarily work- ing with the DFM rule basis. WUS specifical- ly has Joan with multiple applications that, on our roadmap, we'd like to touch in the next three to five years. But some of them are pret- ty far out there. The capability of this technology beyond just straight PCBs is pretty amazing. But we al- ready have the DFM rule basis that we have for applications on GPUs, CPU type technologies, graphic cards, PCIE, mezzanine AI cards, and interposer replacements ATE boards. These are all WUS bases that those markets specifical- ly needed to know how to use it just for that. We've been training customers on how to do that, so they can do the designs. Johnson: The primary area of focus now you have the process underway and—short of hav- ing a place for customers to go for second sup- pliers—it's all about getting designs through to production. Dickson: Our process over the last 12 months is fairly mature. We're looking at doing innova- tions for VeCS for volume. Some of our equip- ment will be changing due to that for lowering costs, getting things more volume specific. But pretty much our process hasn't changed that much in the last 12 months. We're fairly ma- ture on where we are with the technology from a manufacturing standpoint. Now more into material matches. We've run multiple tests on capability for via fill and slot technology. We finally found the end of where if the slot is so deep that the fill material CTE has to match the PCB CTE. But this depth is greater than 20x HDI blind vias, and now we're match- ing the via fill materials with the laminate ma- terials. It's becoming much more sophisticat- ed, and we're tweaking it now so that we can get flat BGA assemblies flatter than most types of assemblies. That's what we think the next big breakthrough is—how to match the system so that we can do really large BGAs. Johnson: It sounds like you are on the verge of a designer's handbook to document how to do this. Dickson: We're getting really close. Matties: Moving into plating, what advice would you give a fabricator generally about the plating process? Dickson: It's a pretty amazing physical phe- nomenon. These slots are very narrow at 0.3- 0.5 millimeters, and they go deeper than 1 mil- limeter, which is a pretty deep blind via. Once you get three times past the width of the slot, the plating distribution acts like a through- hole, even though it's a blind connection. The plating solution travels easily into the slot, and the distribution of the thickness is very uni- form, much more than any other type of plat-