Before mobile operators and OEMs can leverage the potential of 5G’s advantages in data speed, ultra-low latency, and greater capacity, they need to overcome several challenges. These range from cost issues for infrastructure and equipment to the innovation of enabling technologies, including RF filters and solutions, sensors, timing devices, and semiconductors.
One of those innovation companies is Resonant Inc., a provider of RF filter solutions based on its intellectual property (IP) software platform. The company developed its XBAR RF filter technology, which meets the bandwidth requirements for 5G and Wi-Fi 6 and 6e applications, using its legacy finite element modeling (FEM) software platform, called Infinite Synthesized Networks (ISN). The company recently upgraded to the new WaveX design platform, introduced in June, which offers 3D FEM simulation capability to design RF filters, leveraging a hybrid multi-cloud, large-memory GPU-powered implementation.
WaveX encompasses a a suite of proprietary algorithms, software design tools, and network synthesis techniques. It allows the Resonant design team to identify causes of spurs, optimize isolation, minimize insertion loss, manage bandwidth, and center frequency, which are critical requirements for high-performance 5G, Wi-Fi, and Ultra-wideband (UWB) filters.
Resonant provides design simulations, based on surface acoustic wave (SAW) and temperature compensated SAW (TC-SAW) technologies, to its customers which they manufacture in their fabs or at one of Resonant’s foundry partners.
Electronic Products spoke with Resonant’s Mike Eddy, vice president of corporate development, about RF filters as an enabling technology for 5G. Here is an excerpt from the discussion.
Electronic Products: Where are filters used?
Mike Eddy: Filters are used in many different devices but in particular in smartphones where you need to manage multiple frequency bands. A typical smartphone, such as an iPhone 12 or 11, will have somewhere in the order of 60 to 80 filters.
You can’t tune filters, therefore, every different frequency band in a phone needs a filter. And because everything is about bandwidth you also need to aggregate multiple bands, so that’s why a smartphone has many antennas. Each antenna needs a whole new set of RF components which means a whole new set of filters. So, for instance, an iPhone 12, I think had something like five antennas – one was for Wi-Fi Bluetooth and GPS and the other four were all for cellular.
The interesting aspect of 5G is that it is a new set of frequency bands that are at a much higher frequency. 3G was primarily about 1-GHz frequencies; 4G was about 2-GHz frequencies, and as you look at 5G where you want to have very, very fast data rates and very, very short delays, you need wide bandwidths, and the frequencies below 3 GHz are now full. So for 5G in order to get new frequency bands that are wide in bandwidth, which means high data rates and low delay, you needed to look above 3 GHz. That’s why with 5G you hear about these bands in 77 and 79 millimeter wave (mmWave). They’re both above 3 GHz, so for instance n77 is in the 3.3 to 4-GHz range, so it’s a much wider bandwidth but again a much higher frequency.
5G sub-6-GHz and Wi-Fi bands in the 3 GHz to 7 GHz frequencies (Source: Resonant) Click for a larger image.
Electronic Products: How does 5G change the requirements for filters?
Mike Eddy: From 3G to 4G you went from about 1-GHz to 2-GHz frequency, and the bandwidths went from 30 to 40 MHz to 60 to 70 MHz. When you were at 1 GHz, 3G, the acoustic wave of resonating or building blocks of filters was called SAW or surface acoustic wave. It’s a very simple device with a metal dome structure on top of a piezoelectric substrate, and the metal is physically moving in order to generate a resonance from which you build a filter. That was the perfect piezoelectric resonator for that kind of frequency and bandwidth.
When you move from 3G to 4G then the requirements are very different, it’s 2 GHz [frequency] and it’s now 60 to 70-MHz bandwidth. SAW could be pushed in order to meet these new requirements but it was not optimum.
And so Broadcom, Avago at the time, came up with what was the optimal structure for these new requirements, which they called FBAR [film bulk acoustic resonator]. It’s basically a metal drum around a piezoelectric [substrate], which is aluminum nitride, and that metal dome then vibrates at a frequency of about 2 GHz with the right kind of bandwidth. [Avago acquired Broadcom in 2015.]
Now going from 4G to 5G, again, the requirements are very different. You’re talking about 3.5-4.5 GHz and it’s not just 5G. The new Wi-Fi bands are at 6-GHz and ultra-wideband (UWB) used in the car for very accurate location is in the 7-GHz range. It is all driven by the fact that you need these wide bandwidths and there’s nothing available below 3 GHz.
The manufacturers who made out well in 4G, they are trying to extend the performance of their bulk acoustic wave (BAW) structure, the FBAR, by doing materials doping and adding extra elements in order to meet the new requirements for filters. Since we’re a licensing company and have no legacy baggage based on engineering, manufacturing, or foundry, what Resonant did was to look at what would be the best structure, or building block, for these new requirements.
We have a very powerful and accurate software tool and screened hundreds and thousands of different kinds of structures for these filters and came up with XBAR [a type of BAW acoustic resonator], which we believe has the right coupling coefficient in order to meet the requirements of 600-MHz, 900-MHz, 1200-MHz bandwidth at frequencies from 3-13 GHz, and even beyond that. But our main focus now is in the 3-8 GHz range, which is where most of the market is right now.
XBAR resonator, showing cross section of basic structure and bulk wave excited by metal interdigital transducer (IDT) fingers (Source: Resonant) Click for a larger image.
Electronics Products: In addition to new performance requirements, how challenging is it to keep shrinking the size of the filters in order to fit so many of them in a smartphone to handle the different frequency bands?
Mike Eddy: It’s absolutely critical. When you look at how thick a smartphone is these days, you really have to be careful about the height of the filters. A typical bandpass filter in a phone is about one-millimeter square by about .35-mm thick. So these things are just specs of dust but there are so many of them that they take up a lot of area.
Electronics Products: What happens if you don’t select the right RF filter for your application? How does it impact performance?
Mike Eddy: If you look at a teardown of a 5G module and the 5G filters that are being used right now in the early stages of 5G, the filters are very poor. They don’t reject potential interference very well and the reason they can work right now is because there is just not a lot of traffic. As you get traffic on these new 5G frequency bands you need to reject these potential interferers. The key to performance is what’s called the signal-to-interference and noise ratio [SINR], so you want the signal way above the noise floor.
And the more above the noise floor you have that signal then the faster the data rate and the better the user experience. If you can’t reject the potential interferers in neighboring frequency bands, then that interference raises the noise floor. Therefore your signal-to-noise gets worse which means your data rate gets significantly degraded, which ultimately means you can’t get the video quality you want and you get a dramatic degradation of battery life because your battery is trying to work harder in order to get the signal.
The other part of 5G is all about data rate and delay. If you’re spending a lot of time retransmitting because the quality of your signal is not good, then you’re delay goes up too. So that’s why filters are so important because if you really want to get the experience that 5G promises you really want to protect that bandwidth. Verizon paid over $40 billion for the C-block in the U.S. and the last thing they want to do is to pay all that money and find the user experience is no way near what is anticipated for 5G.
Electronics Products: How does SINR degrade battery life?
Mike Eddy: A physics theorem called Shannon’s Law [Shannon-Hartley Capacity Theorem] tells you what kind of capacity and data rate you can get out of an RF channel. It’s a very simple formula: it is bandwidth, the number of RF paths in the phone and the number of antennas, and the signal-to-interference and noise ratio.
These different generations of wireless technology are affecting those three things, so bandwidth means new frequency bands that are wider and aggregating multiple frequency bands; it’s about the number of RF paths which is the number of antennas, and signal-to-interference and noise ratio, which means protection with filters and bringing the signal source closer to the user. That’s why with 5G you hear a lot about densification of the network and small cells because the closer you bring that signal source, the higher the signal strength, so the greater the SINR – the signal-to-interference and noise ratio.
Electronics Products: What is the XBAR technology and how does it solve some of these 5G challenges?
Mike Eddy: Our software tool is a finite element modeling (FEM) tool called Infinite Synthesized Networks to develop complex acoustic wave filters. If we know the materials properties and physical dimensions then we can accurately predict the filter performance. We developed it for surface acoustic wave (SAW), which is a 3G technology, and TC-Saw [temperature compensated SAW], which is an extension of that 3G technology for 4G.
About three years ago, we were looking to extend it for bulk acoustic wave (BAW) technology but we also realized that 5G was coming along. As we looked at 5G we couldn’t see how the 4G BAW technology would work for these new requirements. So rather than spend time on that we decided to look at what would be the best acoustic wave technology for the new requirements of 5G. And that’s when we invented XBAR. With the intersection of our very accurate software design technology, engineered substrates, and our talented design team, we developed XBAR, which is a BAW technology that is very different from anything that existed previously. It matches the wide bandwidth, high frequency, and high-power handling requirements [of 5G].
The other part about high frequency means the signals don’t propagate as far as the lower frequency signals, so the way to get around that is to increase the power. That means all the components in the RF chain need to be able to handle higher power and we found that XBAR could meet all of these new requirements.
We showed the [XBAR-based] 5G RF filters to all the major players in the industry. They were very excited about it because everybody understood that the requirements were changing for 5G. At Mobile World Congress 2019, we showed a demonstration filter using XBAR that had all the right requirements for 5G. It is called the n79 filter with a 600-MHz bandwidth from 4.4 to 5 GHz, which rejected the neighboring n77 and Wi-Fi signals.
The company that moved the quickest to work with us was a Japanese company – the largest filter manufacturer in the world. They very quickly signed an agreement to invest in Resonant and also signed a development agreement for RF filters using XBAR. We’ve been working through that development program and we met the second major milestone late last year. This means we’ve met all of the technology requirements for performance, packaging, and reliability and they’re moving that [XBAR-based RF filters] to commercialization. We expect to see our designs in 5G, Wi-Fi, and UWB as part of the smartphone market as we enter 2022. [Murata signed multi-year commercial agreement for XBAR in 2019.]
Key characteristics of XBAR resonator addressing 5G requirements (Source: Resonant) Click for a larger image.
Electronics Products: Is there anything else you’d like to add for engineers and designers involved in 5G designs?
Mike Eddy: There’s always smart people in any company and there’s so much legacy engineering, IP, foundries, and fabrication that’s associated with all those previous technologies. They will find a way to make it work for 5G – the wide bandwidth and high frequency – but it’s always a compromise.
XBAR can cover the entire bandwidth with a single filter. We’re already seeing what the other companies are doing. They are doing multiple filters to cover the bandwidths. That will work but in an XBAR world it’s always going to be non-competitive because it’s going to be bigger, it’s going to be more costly, and it’s going to have lower performance because you have more losses compared with a single device that covers the entire frequency band.
Everybody knows that in 2022 and 2023, when 5G networks are in place you’re going to need high-performance filters in order to get that 5G experience. 5G as it stands right now is wave one, which is more of a marketing experience than a user experience. It is which carrier can claim the first nationwide network and which carrier can claim the best performance. The performance of 5G right now is about the same as 4G.
Once you get the network deployed with new spectrum bands and new frequencies and you start to densify the network that is when you’ll see 700 megabyte/1 gigabit per second kind of rates with a less than 10-millisecond delay. This is in comparison to the 20-30 megabits per second that you see right now because we haven’t deployed the real network yet.
Electronics Products: Do you think RF filters are a performance roadblock to 5G?
Mike Eddy: For example, a typical movie download takes about 25 minutes right now on a typical 4G or an early stage 5G network. If you have a full 5G implementation with high-performance filtering it’s on the order of 13 or 14 seconds.
5G is very different from the previous cellular generations. It is extending that data speed, the cord cutting, and use [cases] of wireless for your everyday life. The combination of data speed and reduced delay enables new applications. This is why you hear a lot about private networks, Industry 4.0, autonomous cars, and remote surgeries. It’s starting to open up new applications because of the totally different uses that can be envisaged for high data rates and low delay.