Issue link: https://iconnect007.uberflip.com/i/1507822
SEPTEMBER 2023 I PCB007 MAGAZINE 15 and laminates in the computerized systems are required to handle those materials; the cost to enter that technology is prohibitive for some, but not for others. Let's look at glass fabric. In the 1990s, we had generally four glass styles: 7628, 2116, 1080 and 106. Ever ything you produced could use one of those four glass styles. Now, there are many more glass styles in regular use, plus thinner mob i le b oard s. You have 1037, 1027, 1017, and the next wave may end up being 1010 or something even finer than that. Consider that the weavers now must handle a finer filament. You can't weave as fast, and those finer glass styles also have more weave per inch. You're tying up that equipment for a longer amount of time to get the same square meters of mate- rial. erefore, it will cost more from a capi- tal utilization. ere are many things that will challenge the cost. e material suppliers have to balance those challenges. Johnson: Is the demand for UHDI-type manufacturing load increased? Demand is definitely increasing. Holden: What is Panasonic working on with respect to a non-clad, additive build-up film (ABF)? e challenge is to be able to direct metallize a dielectric without the copper falling off. e other challenge that the whole industry runs into is there are many different resin systems out there. Naturally, each supplier is trying to protect their intellectual property. ere are tons of patents, so it becomes increasingly dif- ficult to navigate through the patent landscape. is is because in the PCB space, they're still modified with some additive plating or tradi- tional buildup type, some manufacturing of circuitry. So, will it eventually go toward fully additive? As they get smaller, I'm not sure. It gets harder to find the ability to use a thin start- er foil, because right now on mSAP it's usually somewhere between 1- and 3-micron foil. If you think about PCB etching, typi- cally it's a one-to-one aspect ratio. If you are going through 18-mi- cron (half-ounce) copper, it's hard to get smaller than an 18-micron line. So, if you translate that into ul- tra HDI where you're look- ing at, let's say, a 5-micron line, 3 microns of copper foil begs the question: What's le over to plate your via? Even the ability to have a foil there is increasingly challenging. Johnson: Must you be a greenfield facility to do this sort of work, or can you grow into it? In Asia, the facilities that introduced mSAP technology were doing other technology at the time that it was introduced, so they showed that it doesn't have to be a greenfield facility. But if you look at the volume of handheld-type devices in the world today, that has spawned a significant number of mSAP lines. To enter that space takes a huge capital expenditure, not only in your cleanrooms and conveyorized equipment, but in plating technology. Green- field might be the easiest way to do it, but that doesn't mean an existing facility can't do it. Matties: Is the market space one that if you build it, they will come, or do you go out, develop customers, and then invest? It might be good to have a customer in mind beforehand because it's a huge investment. Just because someone has that equipment doesn't mean a customer isn't willing to try. Now, if you have an OEM that needs that tech- nology, are they willing to take a chance on somebody new, or do they just want to invest in somebody who already knows their products? The challenge is to be able to direct metallize a dielectric without the copper falling off.