Reading time ( words)
At the 2018 electronica exhibition in Munich, John Hendricks, product marketing manager for Rogers Corporation, discussed 5G materials including demands and trends.
Pete Starkey: John, it's great to see you again. Thank you for sparing the time to join us. It's been a very active show, and everybody I've spoken to has been very complimentary about it. One of the main talking points this week has been 5G. From your point of view, as a specialist supplier of materials, where are we going with 5G, and what are the implications of 5G for your company?
John Hendricks: Rogers is a company that makes printed circuit board materials, but most specifically, high-frequency PCB materials. We tend to approach things from the high-frequency point of view. If you look back at previous generations of telecommunications systems—2G, 3G, 4G—there were many technological developments along that path and evolution. In terms of high-frequency PCB materials, not much has changed. A power amplifier was a power amplifier, and an antenna was an antenna. The materials used in 2G also tended to get used in 3G and 4G. Also, the frequencies were very similar.
However, 5G is different from that point of view. There are a couple of reasons for that. One is there is a split in frequency. There's a low-frequency bit often called sub-6 GHz, and there's a high-frequency part generally called millimeter wave technology. The millimeter wave starts at 28 or 39 GHz, and the requirements for a PCB material at those frequencies are different from what they are at 1 or 2 GHz. That's one completely different change.
Secondly, even in the lower frequency area—sub-6 GHz, which is typically 2.6, 3.5, or 4.8 GHz—there's a trend toward much more integration. A power amplifier in the past was a single component. Now, power amplifiers, transceivers, and high-speed digital boards are all being integrated into high-count multilayer boards, typically 10–20 layers. That presents significantly different challenges for the materials than a two-layer power amp or a four-layer board that was used in previous generations, for example.
Starkey: How does your current range of products meet the demands of 5G?
Hendricks: Some of the materials that were used in the past can still be used. We used to break it down, and I suppose we still do break it down into power amplifiers and antennas. The power amplifiers in the past used RO4350B from Rogers, and that still has a strong place in 5G. The antennas used to use a lot of woven-glass PTFE materials typically acquired through Arlon. As we go into 5G, the antennas start to become a little more integrated. You might have a four-layer multilayer board or six-layer antennas with built-in feeds and distribution networks and elements.
That tends to move towards the more thermoset materials like RO4730G3, so that's covered. We've been supplying into millimeter-wave PCBs for many years, typically 24 and 77 GHz for automotive radar sensors; we have materials like RO3003 for that. Our traditional materials have a number of bases covered, but what we didn't really have was materials that were suitable for low-loss RF high-count multilayer boards, which is that integration that I referred to earlier—the power amplifier, transceiver, and high-speed digital coming together.
It's still ongoing, but what we did about six months ago is launch a family called RO4000T, which is a big extension of the RO4000 family. Essentially, those materials are thinner with many more options. Instead of having nothing below five mils or whatever it might've been, we now have 2.5-, 3-, 4-, and 5- mil thicknesses; it's a complete family of laminates, prepregs, and low-profile copper foils. The differences are that the new materials are thinner, but they also have spread or flat glass for reduced skew or variation in the dielectric constant.
Starkey: What types of resins are involved?
Hendricks: It's still the same thermoset resins. The RO4000 resin system is basically the same, but we're using smaller particle fillers, flatter glass, smoother copper, and more options. When you look at those, they're called RO4835 laminate, RO4450T prepreg, and CU4000 copper. The biggest difference is that in the past, people used to do a lot of core lamination with our products, and other people did foil lamination elsewhere. With these, you can do three- to four-stage HDI with three- to four-stage sequential foil laminations, using these copper foils as well as the prepregs, which is quite new for Rogers materials.
It's also a challenge because the higher frequencies need smoother coppers, which tends to mean poorer peel strength or less reliable PCBs. What we found over the last six months is that people who test these new products are reporting that even with three- or four-stage sequential foil laminations and with smooth foils, you're still getting good copper bonds and highly reliable bonds and plated-through holes through multiple thermal cycles, for example. I think people always expect the Rogers materials to have the best performance, but it's also proving to be extremely reliable through complex manufacturing processes. And that's probably the most important thing, which is so encouraging.
Starkey: Is there anything else related to materials for 5G applications that would be interesting to talk about?
Hendricks: If you look at 5G, one of the things people say—some of it may be a bit of hype, and some of it may be real—is that there are multiple “use cases” in addition to just phone calls and downloading videos. They tend to talk about three spaces: enhanced mobile broadband, which is essentially mobile or fixed communication; massive machine-to-machine communications, which is basically the next generation of IoT and is coming along a little bit later; and ultra-low latency, high-reliability applications like using 5G in autonomous driving, which is obviously something for the future but not today.
If you look at the immediate application, that's mostly the enhanced mobile broadband. In the millimeter-wave area, that typically means fixed wireless access at 28 or 39 GHz at the beginning; in the U.S., it may be an alternative to cable technology. What's coming in the next one to four years is the use of millimeter-wave bands for mobile communication. People are developing 28-GHz handsets, and that's something that we've never had in the industry before.
That’s a challenge for the materials because in those handsets, materials need to be not only low loss with smooth copper, they also have to be compatible with these multilayer manufacturing processes, for example. They sometimes need to be flexible and halogen-free because halogen-free is necessary in handsets but not in base-station infrastructure. That's a big challenge, which is something where we have some liquid-crystal polymer (LCP) materials that partly get you there, but it's an area that we're looking at closely—the handset people are actually telling us that they need our materials, not just the base-station people. That's one of the more interesting thing going forward for us.
Starkey: If you have the right knowledge and forward vision, you can anticipate what the requirement is going to be and be prepared when people ask for it.
Starkey: John, I've learned a lot in the last few minutes. Thank you very much for being so open and sharing your information. I really appreciate it, and it's great to meet you again.
Hendricks: Cheers, thanks.