Reprinted with permission from Evaluation Engineering
Among the relatively overlooked aspects of our modern wireless society we often bring up is the aspect of infrastructure, both wired and wireless, that keeps it all going. Our shiny functional world is supported by a massive amount of wiring behind the wainscoting. Proper use of the RF spectrum demands the best of every aspect involved, from the software to the antennas.
To get a deeper look at one of the aspects of the Cloud and managing the RF spectrum, we reached out to Michael Eddy, VP of Corporate Development at Resonant. The company has been working on solutions to address spectrum deficit, which can mean reduced data speeds and a reduction in usable bandwidth. RF filters are one of the ways to enhance spectral efficiency by protecting from degradation due to interference.
How efficiently a given frequency carries a data packet is critical to maximizing a system’s RF performance. Filters with wide bandwidth are needed to realize the full potential of 5G and Wi-Fi. Resonant’s XBAR resonator can create filters that deliver performance and maximize efficiency at up to 1,200MHz, with support for frequencies higher than 3GHz, with very high power handling and low loss.
EE: Thinking about how far RF has come as a application space, I mean, 50 years ago, you could count direct RF applications on one hand. After you get past consumer, there were a few military applications, a handful of industrial applications. What would you say is the big defining line in the RF space, where RF stopped being a niche application space as it were, and really started becoming a commodity for society?
Michael Eddy: I think the driver and the inflection point happened because of the phone, the wireless phone. At the very beginning, and I remember, if you see on the TV, these very early Motorola phones, which looked more like a brick than what you see in a modern smartphone, the key was the RF. But no one really knew how that was going to take off, as you started to really engineer that RF component in the phone, so that you could make it both mobile and have distinct functionality, and allow you to become wireless as opposed to wired.
EE: You could make the argument then Mike, it's almost like an iPod, iTunes thing, that the smartphone is now ubiquitous, a mainstream RF device, but that it also required the connectivity aspect to be realized to make it a thing. It's one thing to say, "Okay, everybody can get a walkie-talkie," but then to say, "everybody can now get a walkie-talkie that is indistinguishable from a telephone in your mind."
Michael Eddy: Initially it was just the voice and putting that code on the phone, but once you started integrating it with the internet, it really took off as a device that was essential for everybody, both to make calls and also to interface with the internet and become more and more sophisticated as each generation of wireless technologies evolves and becomes more and more sophisticated and more and more functionality.
EE: We see multiple issues when we start doing what we call the cloud, right? You've got multiple devices, using multiple bandwidth, with multiple protocols. I mean, there are a lot of moving parts, no pun intended, out there. What are some of the issues you see?
Michael Eddy: Really, the issues we see are more of a function of the complexity and the amount of signals being transported from your phone to the network, and from all the different devices to the network and then the internet, and being able to define and maintain the lanes of the different signals, to make sure that they reached the place they need to go to, with high integrity.
EE: What about the people who would say, "Well, I've got software defined radio now, I've got digital devices, I can just define in the software how exact I want my signal to be."
Michael Eddy: You can do that to an extent, but the size and the amount of these signals now are such that you're still required to deal with that integrity and maintaining the lanes, the frequencies of these signals, using RF. It's only through, by very sophisticated filtering, allowing the signals that you want to listen to and to deal with in the phone and vice versa, that you can maintain the integrity of those signals. That's why high-performance filters in the phone have become an essential piece of maintaining and making sure that the performance of the phone is maximized.
EE: What kind of performance on these filters are we talking about? Are these straight up brick wall filters or is there an attenuation roll off? How precise are the current filters and how precise do they need to become?
Michael Eddy: Just as a stake in the ground, take a modern smartphone. A high-end smartphone has somewhere between 60 to 100 filters in there. That's because these phones need to operate on multiple frequency bands, anywhere from 600 megahertz and now in 5G and Wi-Fi, up to seven gigahertz. Each particular band that you want to use in that phone needs a filter, so that the different signals do not interfere with the signal that you're interested in.
The key to these filters are they have to have very low loss, so that you can maintain low power so that your battery is not drained. That also gets the maximum signal strength, so you can get the highest data rates possible out of your phone. Then, it must be rejecting any of the out-of-band signals that are potential interference. So, as you said, the roll-off needs to be very steep. The closer you can get to a brick-wall filter the better, and that involves what we call acoustic-wave filters. These filters are based around the piezoelectric effect and the roll off you can get in this very, very small size is incredible. So, you're typically attenuating a potential interfering signal by somewhere in the order of 50 DB.
EE: That's very interesting Mike, because the power space is being disrupted right now by GaN, which is a piezo semiconductor. It's interesting that piezo materials, which used to be a novelty, are now becoming critical parts of our infrastructure.
Michael Eddy: Using piezoelectric filters was a critical turning point in the phone, because there are so many filters and they need to be so small, that the only kind of filter technology that you can really utilize with this kind of performance and this kind of size, are piezoelectric filters. They're basically using the piezoelectric effect to generate resonances that can be built up as a filter. As you say, it's a very interesting time right now because 4G is transitioning to 5G, and obviously when that happens, there are some key disruptions going on because of the change in requirements.
You mentioned GaN, which is a very, very good power semiconductor for high power and high frequency. That's why you're seeing GaN now being used for infrastructure for 5G, because 5G is now moving to higher frequencies and a much larger bandwidth. That's exactly what's happening, so you'll see the replacement with GaN on the power side. Also, what you're seeing now in the phone is a move to different kinds of filters, still based on the piezoelectric effect, that can optimize the performance because they work best at high frequencies and wide bandwidth.
EE: Some of our audience knows that my background is Army electronic warfare, which was in my day was nothing but old-school RF and we were almost out there in the woods with coat hangers. To see the migration now to such high levels of precision, it's really something. For example, if you wanted to look at another application space that's completely removed from ours, but is actually getting a weird renaissance because of accuracy of precision is ancient archeology. We now have the tools to appreciate how precise that stuff was. 18th- and 19th-century man didn't have the tools to know how precise that stuff was.
Michael Eddy: That's a really great analogy and perhaps I can just build a little bit on that, about what Resonant does and our technology for 5G, which we call XBAR. In cellular wireless technologies, we go by different generations. So, 3G technology, the frequency bands were about one gigahertz with a bandwidth of about 35 megahertz, and a certain kind of acoustic-wave filter technology is optimized for that particular set of requirements. And that's what's called SAW, surface acoustic wave, acoustic technology. When 3G went to 4G, you're now looking at about two gigahertz frequency bands, with a wider bandwidth of somewhere in the order of 65 megahertz. So, the bandwidth increased, and the frequency increased, and the old technology for 3G SAW was really not optimized for this new set of requirements, and a new acoustic wave resonator technology was developed for that particular new set of requirements, which was called FBAR.
It was actually invented by Broadcom, then Avago, and that is a bulk acoustic-wave technology that was optimized for these new requirements. 5G is again, very different. You're now talking above three gigahertz with bandwidths of 600 megahertz, 900 megahertz. The new Wi-Fi bands at 1,200 megahertz wide at six gigahertz. These new requirements, the old 4G technology, there is a huge amount of fab infrastructure, engineering infrastructure around that 4G technology.
Those companies are trying to extend the performance to meet these new requirements. But Resonant is an IP licensing company with some very sophisticated software, so what we did is we took a blank sheet and said, "What is the best technology for these new requirements?" So, we invented this new resonator technology for filters, optimized for these new requirements and we call it XBAR. Our XBAR technology is designed for these 600 megahertz, 900 megahertz, 1,200 megahertz bandwidth, at frequencies of three to seven or higher gigahertz. Does that make sense?
EE: It makes a lot of sense, and actually, I wanted to comment on what you said and follow up, because it is so fascinating to hear the development path for Resonant's solutions. You could tell that it's an example of technology convergence in that you're a material scientist, a professional, and you've come up through these spaces because you've touched the spaces by adding value with material science, and it's almost a chemical engineer's approach. "Okay, what's the shape of the molecule? What molecules can I find that'll have that shape and let me make this new drug." It's interesting to see the different approaches refreshingly creating these enabling solutions.
Michael Eddy: Actually, it is a tremendous convergence of some really critical technology. One of which, as you said, is the material science, and the material science that has been critical is what's called engineered substrates. The ability to put different layers of single-crystal substrates on silicon. So, we can now put things like lithium niobate and lithium tantalate, these piezoelectrics on silicon. This is a very recent establishment of what we call engineered substrates.
Then the other convergence within Resonant, is we have this very sophisticated software platform, it's called a full finite element modeling tool, where when we have the materials, properties, and physical dimensions, we can accurately predict the kind of RF performance we will get out of a fabricated structure. I think we're the only ones who have that level of predictability and accuracy around our software tool. So, the convergence of the materials and the sophisticated software allowed us to invent the right kind of structure for these new requirements.
EE: Well, kudos. One of the ways to determine a technology's sophistication, is how precisely can they apply force. From military weaponry to RF signaling, the thinner your scalpel blade is and the more precisely you can place it, the better a doctor you are, as it were.
Michael Eddy: Absolutely. We're firm believers in that, that we're continually improving our software tool for accuracy because that's what really allows us to invent and really make sure that we are at the forefront of the kind of technology for this next generation wireless. And it's not just 5G, all of the frequency bands below three gigahertz are full, so all of the new wireless technologies are above three gigahertz. That's why 5G is in the 3.3 to five gigahertz range. The new Wi-Fi bands are five to 7.1 gigahertz, ands ultra wideband is being used now in the iPhone 11 and 12, for very precise location. That's in the six-and-a-half to nine gigahertz range. You'll see even Satcom looking in the 12 gigahertz range, that's where the bandwidth are available. We believe our technology will be able to utilize and take advantage of all of these new applications.
EE: You touched on one of them, when you said location, smartphones and personal electronic devices tied to the cloud are going to become a critical part of sensor fusion in a smart city environment, intelligent traffic and the like. Because the car may see you, but then if the car also sees the phone as it were, you're doubly safe.
Michael Eddy: Yep. If you look at a modern automobile and you look at the shark fin, that is a very sophisticated. It's called a telematics control unit, a TCU, and it's building on the kind of modules you see in the phone, because it needs to interface to the network. Then, it needs to interface within the car and interface with all of the sophisticated sensors that are within the car. And so already, you're seeing a very sophisticated wireless module that is being used in the car, as we move towards more and more autonomy and more and more need for infotainment within that car. And that's why you see rumors of people like Apple looking at autonomous vehicles, because everybody is fighting for that, what they call a full screen, the screen within the car.
EE: The car is literally the convergence of every technology currently being developed. Everything from wide-bandgap semiconductors to advanced RF filtering, it's going in the car.
Michael Eddy: And as you said, as you move to more and more autonomy, it becomes very, very critical. Part of 5G is very wide bandwidth. The other part is very, very short delay times. Latency getting less and less, and that's going to be critical in the car, and if you think of safety, moving at 60 mile an hour, if you have something that is going to have a long delay, that is going to be a problem. And so again, another key piece of 5G is to move the compute piece to the edge, closer to the user, so that the delay becomes minimal.
EE: That's the whole impetus behind, for example, artificial intelligence at the edge. Everything that you can do to leverage computing at the edge is being looked at because latency is so critical.
Michael Eddy: Yep. No, I'm very excited about the transition from 4G to 5G because it is such a disruption and it's very different from 4G because it's going to enable new applications beyond just the wide bandwidth. Whether it's more autonomy in the car, whether it's more robotics in the manufacturing line, whether it's increasing use of AR and VR, because the latency's so short now. There's just a raft of new applications that can be generated because of these very different and improved requirements of 5G.
EE: Excellent. If you had any final thoughts, now would be a perfect time for them.
Michael Eddy: I think basically, making sure that people understand that we're in the early stages of 5G. Each generation in wireless lasts about 30 years, and each generation of wireless starts at a clip of about every 10 years, and we get a new generation. Right now, we're at the early stages, and so you're not seeing the dramatic improvement over 4G yet, but in the next two, three, four, five years, you'll start to see some great improvements in the data rates for wireless and 5G.
You'll also start to see the need for these very high-performance filters, that we are at the front edge of, at Resonant, generating our XBAR filters. And the validation beyond, you're absolutely describing this correctly, is we have the very sharpest scalpel with our software tools, is that the largest filter manufacturer in the mobile space is partnering with Resonant to bring this technology to the marketplace.
