Just as solid state replaced and revolutionized the recording and storing of audio, video, and data, silicon-based speakers could soon replace mechanical coil and magnets in true wireless speakers (TWS) – earbuds and headphones – and even hearing aids. Solid-state speakers could lower earphone production costs, reduce product variability, reduce returns – and produce higher-quality audio in even lower-priced models.
xMEMS Labs, a Silicon Valley company founded in 2018, announced it has started shipping two solid-state micro speakers: Cowell, what the company claims is the world’s smallest solid-state micro speaker for TWS earphones and hearing aids, and the Montara Plus, “the world’s highest output MEMS micro speaker,” capable of SPLs up to 120 dB at 200Hz, and designed for high-resolution in-ear monitors. With either model, the result is a high-performing speaker about a tenth the weight and 40% of the size of a similar coil speaker, even meets IP58 water and dust protection levels, according to xMEMS.
The company also announced it has started to ship its Skyline DynamicVent, what xMEMS claims is the world’s first piezoMEMS, DSP-controllable vent to reduce unwanted earphone and hearing aid occlusion effects.
“What we are doing is attempting to reinvent the coil speaker, the coil, magnet, and paper diaphragm speaker, a mechanical speaker, which has been around for 100 years, into a semiconductor speaker,” explains Mike Housholder, xMEMS VP of marketing and business development. “With that transition, it will become easier to manufacture [speakers] at high volume, it will become more reliable, it will exhibit better performance, better audio quality.”
According to Housholder, xMEMS new silicon speaker architecture delivers 150 times faster impulse response than traditional designs compared to the slower attack and decay of coil architectures. “What you get is some of the performance benefits of a ribbon planar magnetic speaker in terms of better acoustics, and better speed,” Housholder says, “but you’re getting something that is highly manufacturable, highly reliable, and highly scalable.”
Technically, the coil on magnet is replaced with a piezo layer. A thin film piezo layer becomes the actuator, which bends when voltage is applied, and silicon –95% stiffer than paper or plastic – serves as the speaker diaphragm. As the piezo material bends, it pushes on the silicon, which then moves air to generate sound. “You have no speaker breakup, fast impulse response, and no phase shift,” Housholder exudes. “This is a voltage-driven speaker technology versus capacitive, so the transient response, the impulse response of the speaker, is instantaneous. That combines to deliver fantastic sound. The mids, the highs, the detail, the quality, the resolution is next level.”
Housholder says that every major audio brand is currently experimenting with samples and building prototypes. Q3 would be the earliest manufacturers could begin production, but it may “be many more months until the silicon speaker finds its way into an end consumer product.”
Easing production, lowering costs
For earbud and headphone makers, solid-state speakers will likely be “easier to manufacture, more reliable, better quality, and in general deliver better performance than their non-solid-state variants,” Housholder notes, especially since sourcing magnets have become more challenging and their performance more variable. Manufacturing of current mechanical speakers “is not fully automated. You’ve got to have operators on assembly lines wrapping copper coil around magnets, attaching them to mechanical framing, and then putting a paper or plastic diaphragm on top. Any time humans are involved you have part-to-part variability. Humans involved create variability.
“You move this all to a solid-state semiconductor speaker, you solve a lot of these problems,” Housholder continues. “A semiconductor speaker is a fully automated assembly line, from wafer fabrication to packaging to test, fully automated, much less labor to produce a speaker, much more part to part consistency, speaker to speaker.”
Manufacturing efficiencies, increased solid-state reliability, lower assembly line waste, less left-right calibration matching and testing, and fewer consumer returns could make up for initial solid-state speaker price premiums. But, notes Housholder, “As you move on in time, as wafer costs reduce, as you get into higher volumes, that cost delta shrinks compared to the incumbent technology.” Since solid-state speakers will be price competitive with or, eventually, even less expensive than current mechanical speakers, Housholder expects that “you could find us very easily in a $150 product.”
At the moment, all of xMEMS solid state speakers are manufactured at the TSMC foundry in Taiwan, but Housholder expects to have a completely geographically diversified supply chain by the end of the year, with multiple fabs and multiple packagers. Housholder also notes that while xMEMS’ technology is patented, “we certainly expect other manufacturers to come online, so there will be other options out there in the future,” furthering the potential of a solid-state headphone speaker transition. Housholder expects a solid state earphone transition similar to that of the microphone; according to SAR Insight, in 2007 MEMS-based microphones accounted for just 5% of the total market but 80% of all microphones by last year.
In addition to Cowell and Mantara Plus for TWS earphones, xMEMS also has started shipping its complementary Skyline DynamicVent, which enables control over occlusion effects in headphones and hearing aids. “Noise cancellation today and these transparency modes are all digital,” Housholder expounds. “And with that comes artifacts. If you can put a Skyline DynamicVent over a hole in the earbud, you now have the ability to shut that door or open that door, based on your environment. If you’re out for a run, you want to open that vent to get footsteps out of your head, but also to be aware of your environment. Close that vent and block out wind noise. If there’s a car coming, you want to be able to hear that car but still enjoy your music.”
As a result, the DynamicVent solves the open vs. closed earphone debate. An earphone’s processor can examine inputs from all microphones or sensors, then can open or close the vent according to conditions and situations.
According to Housholder, the DynamicVent, combined with the Cowell speaker, also presents solutions to several occlusion problems presented by new closed, canal-sealing Bluetooth OTC hearing aids, such as the ability of a wearer to hear their own voice more clearly, reduce both the amplification of chewing noises and feedback, as well as standard voice-only hearing aids.
For instance, Cowell, combined with the DynamicVent, replaces the basic hearing aid balanced armature. “The reason [balanced armatures are] in every hearing aid is because they’re small and low power,” Housholder says. “But now hearing aids are getting Bluetooth connected and you no longer want to just amplify voice. You want to amplify media. You want to listen to your music. You want to listen to movies. And that requires a full bandwidth speaker. But these balanced armatures don’t deliver full bandwidth. They’re only tuned to deliver voice. Cowell is a true full bandwidth transducer the size of a balanced armature. You can insert this deep in the ear canal like an RIC hearing aid and get full bandwidth audio. So that’s a game changer for hearing aids.”
See also: LG Electronics Announces Availability Of New XBOOM Portable Tower Speaker