# How Chips Are Using On-Chip Acoustic Resonators for RF Filtering

An episode of [The Hardware Podcast with Fexingo: Chips, Devices, and Electronics Engineering Conversations](https://invisiblenetwork.ai/podcast/1896799982), 23 Jul 2026, 10 min.

Audio: https://audio.fexingo.com/business/the-hardware-podcast/episode-0128.mp3

In this episode, Lucas and Luna explore how on-chip acoustic resonators are transforming radio frequency filtering in modern electronics. They dive into the mechanics of bulk acoustic wave and surface acoustic wave resonators, explaining how these tiny components enable the multi-band connectivity in your smartphone. The discussion covers the shift from discrete filters to integrated solutions, with a focus on how companies like Qualcomm and Skyworks are leveraging thin-film piezoelectric materials to shrink filter size while improving performance. Lucas breaks down the trade-offs between FBAR and SAW technologies, while Luna questions the durability of these resonators in 5G mmWave applications. They also touch on the role of acoustic resonators in enabling carrier aggregation and reducing signal interference. By the end, you'll understand why these 'invisible' chips are critical for the wireless world and how they might evolve for 6G. #OnChipAcousticResonators #RFFiltering #BulkAcousticWave #SurfaceAcousticWave #FBAR #SAW #5G #Wireless #Piezoelectric #ThinFilm #Qualcomm #Skyworks #CarrierAggregation #mmWave #6G #Technology #FexingoBusiness #BusinessPodcast Keep every episode free: buymeacoffee.com/fexingo

## Transcript

Lucas: Your phone probably handles a dozen different frequency bands at once — LTE, 5G, Wi-Fi, Bluetooth — and the unsung hero that keeps all those signals from stepping on each other is a tiny chip called an acoustic resonator.
Luna: Acoustic resonator — that sounds like something from a physics lab, not inside my pocket.
Lucas: It really is a physics trick. These chips use mechanical vibrations — sound waves — to filter out unwanted frequencies. Think of it like a tuning fork that only vibrates at one specific pitch. If the incoming signal matches that pitch, it passes through; everything else gets blocked.
Luna: So it's a mechanical band-pass filter, but on a chip.
Lucas: Exactly. And the two main flavors are SAW — surface acoustic wave — and BAW — bulk acoustic wave. SAW devices were the workhorses for 2G and 3G. They launch the wave along the surface of a piezoelectric crystal like quartz. But as frequencies climbed into the gigahertz range, SAW filters started to lose efficiency.
Luna: Because the surface waves get too small to fabricate precisely?
Lucas: Partly. The electrode fingers have to be spaced at fractions of the wavelength, and at 2.6 gigahertz those fingers are just a few hundred nanometers apart. That pushes the limits of lithography. Plus, SAW filters are temperature-sensitive — the substrate expands and the frequency drifts. That's a problem when you're trying to squeeze more bands into the same phone.
Luna: So BAW came along to solve that?
Lucas: Right. BAW resonators trap the acoustic wave inside a thin film of piezoelectric material — typically aluminum nitride — sandwiched between two metal electrodes. The wave bounces vertically through the film, not along the surface. That makes the resonator much smaller and more efficient at higher frequencies, and it handles temperature swings better.
Luna: And by 'thin film' we're talking microns thick?
Lucas: Single-digit microns. The resonant frequency is determined by the thickness of that film. For a 2.6 gigahertz filter, the aluminum nitride layer is about two microns. That's ten times thinner than a human hair. And you have to deposit it with atomic-level uniformity, or the filter's center frequency shifts.
Luna: That must be incredibly hard to manufacture at scale.
Lucas: It is. That's why BAW filters were initially expensive — only a few players like Avago, now part of Broadcom, had the process control. But over the last decade, companies like Qualcomm and Skyworks have invested heavily. Today, a flagship phone might have a dozen BAW filters in its front-end module, each tuned to a specific 5G band.
Luna: And this is all to enable carrier aggregation — where your phone bonds multiple bands together for faster data.
Lucas: Exactly. Without sharp filters, the signals from band one and band two would interfere. The filter's job is to pass one band and reject the adjacent one by 50 decibels or more. That's like hearing a whisper in a rock concert. BAW resonators can do that in a package smaller than a grain of rice.
Luna: So what's the next frontier? mmWave for 5G and eventually 6G?
Lucas: That's where it gets interesting. mmWave — 24 gigahertz and up — requires even thinner films, and at those frequencies the acoustic losses start to climb. Some researchers are looking at scandium-doped aluminum nitride, which has a stronger piezoelectric coupling, meaning you can get more bandwidth. Others are exploring lithium niobate thin films.
Luna: How do those compare?
Lucas: Lithium niobate has very high coupling, but it's harder to integrate with silicon. You can't just deposit it in a standard fab. So there's a trade-off between performance and manufacturability. The industry is still figuring out which material wins for 6G, which could push into the 100-plus gigahertz range.
Luna: Speaking of the industry — you know, one reason we can dive into these details is that this podcast stays ad-free and listener-supported. A few folks chip in monthly through buy me a coffee dot com slash fexingo, and that literally covers the costs of making this many episodes.
Lucas: Yeah, it's a small group, but it makes a huge difference. It lets us pick topics like this without worrying about ad inventory or sponsor demands.
Luna: Exactly. So if today's tech conversation gave you something usable, that's the channel. Now back to those resonators — you mentioned that BAW filters are already in most 5G phones. What about the durability? Do these mechanical resonators wear out?
Lucas: Great question. Unlike a tuning fork, which can fatigue over millions of cycles, these resonators vibrate at gigahertz frequencies — billions of cycles per second. But the amplitude is tiny, maybe a few nanometers. So the mechanical stress is low. Most BAW filters are rated for decades of continuous use. The bigger concern is temperature and humidity, but they're hermetically sealed.
Luna: So they're essentially permanent.
Lucas: For the life of the device, yes. The failure mode is more likely in the packaging or the solder joints. The resonator itself is incredibly robust. That's one reason the technology has displaced ceramic and SAW filters in many high-end applications.
Luna: What about integration? Are we seeing these filters built directly into the modem chip or the power amplifier?
Lucas: Partially. Some companies are working on monolithic integration — depositing the piezoelectric film directly on top of a CMOS die. That would eliminate the need for separate filter packages and shorten the interconnect. But the thermal budget of the CMOS process limits the deposition temperature for aluminum nitride. So far, the best performance comes from dedicated filter dies that are then stacked or co-packaged.
Luna: So it's a heterogeneous integration story, like chiplets.
Lucas: Exactly. In fact, Qualcomm's latest front-end modules use a multi-chip module approach, with BAW filters, power amplifiers, and switches all in one package. That's how you get the size reduction needed for phones with 30-plus frequency bands.
Luna: Thirty bands? That's a lot of filters.
Lucas: It is. A typical iPhone or Galaxy might have 20 to 30 filters in its front end. Each one is tuned to a specific band: band 1, band 3, band 7, band 20, and so on. And they all have to coexist without interfering. That's the magic of acoustic resonator design — the filters are so sharp that they can separate bands that are only a few megahertz apart.
Luna: So what's the limitation? Why can't we just keep adding filters?
Lucas: Physical space, mainly. Each filter takes up area on the die. And as you add more filters, the insertion loss — the signal power lost as it passes through the filter — accumulates. So there's a trade-off between selectivity and signal strength. That's where tunable filters come in. Instead of a separate fixed filter for each band, you have one filter whose frequency can be adjusted electronically.
Luna: Tunable BAW filters?
Lucas: Yes. Researchers are using varactors — voltage-variable capacitors — or switched capacitor banks to shift the resonant frequency slightly. It's not easy, because you don't want to degrade the quality factor. But some companies, like Qorvo and Skyworks, have demonstrated tunable BAW filters that can cover multiple bands. That could reduce the total filter count in future phones.
Luna: And for 6G, what's the prediction? Will acoustic resonators even work at 100 gigahertz?
Lucas: That's an open question. The acoustic wavelength at 100 gigahertz in aluminum nitride would be about 50 nanometers. That means the piezoelectric film has to be just 50 nanometers thick — maybe 200 atomic layers. Depositing that uniformly across a wafer is extremely challenging. And the acoustic losses increase with frequency. Some researchers think we'll hit a wall around 30 to 40 gigahertz for BAW.
Luna: So what's the alternative?
Lucas: For very high frequencies, people are looking at bulk acoustic wave resonators using single-crystal lithium niobate or lithium tantalate. Those materials have much higher acoustic velocity, so for a given frequency, the structure is thicker and easier to fabricate. There's also interest in acoustic metamaterials — engineered structures that can manipulate sound waves in ways natural materials can't.
Luna: That sounds exotic.
Lucas: It is, but it's also early stage. For the next few years, the workhorse will remain aluminum nitride BAW filters, gradually extending to higher frequencies with better process control. And the market is huge — every connected device needs filtering. Smartphones, IoT, satellites, even cars for V2X communication.
Luna: So the humble acoustic resonator, vibrating away billions of times a second, is quietly enabling the wireless world.
Lucas: Exactly. It's one of those components you never see, but without it, your phone would be a brick. And as we push toward 6G, the pressure to make these things smaller, sharper, and more tunable is only going to increase.

Shows are listed from Apple Podcasts' charts and directory, and Podcast Index where it is used. Every show and episode belongs to its publisher and plays from the publisher's own feed; this site hosts none of them.