xMEMS preps sound-from-ultrasound MEMS microspeaker
Cette publication existe aussi en Français
xMEMS Labs Inc. has designed a sound-from-ultrasound MEMS speaker that will allow the replacement of coil speakers in noise cancelling earbuds.
The use of an ultrasound carrier and demodulation allows the separation of the achievable sound pressure level (SPL) from membrane displacement allowing for smaller MEMS microspeakers that can provide the bass response required, according to Mike Housholder, vice president of marketing and business development.
The Cypress MEMS speaker achieves more than 140dB @ 20Hz SPL. Allowing for 20dB losses due to earbud venting, this is still loud enough to compete with magnetic coil speakers but with better audio precision across the spectrum.
It means that in future full featured ear buds with noise cancelling can be based on a single MEMS speaker per ear. At present solutions tend to use a two speaker solution with MEMS tweeters for mids and highs and magnetic moving coil woofer to provide the bass response.
The Cypress MEMS speaker will be demonstrated at CES 2024 by appointment only with mass production due late in 2024 and appearing in products in 2025. It is a part of two-chip solution aimed at true wireless (TWS) earphones and other emerging personal audio electronics, including smart glasses and sleep buds.
“By shifting to a sound-from-ultrasound principle, the xMEMS Cypress microspeaker can now officially replace traditional coil-and-magnet speakers in active noise canceling earbuds,” said Housholder, in a statement. “Cypress maintains all of the benefits of xMEMS’ existing speakers, while being 40x louder in low frequencies, achieving a key requirement for ANC earbuds.”
xMEMS is not the only company to be exploring sound-from-ultrasound (see Startup uses MEMS ultrasound to improve audio speaker).
Cypress comprises: a modulator to generate an amplitude-modulated ultrasonic carrier wave and a demodulator signal to synchronously demodulate the ultrasonic wave, transferring the acoustic energy down to the baseband, producing the intended audible sound as a result.
In addition the speed of response on silicon MEMS membranes is superior to magnetic coil driven diaphragms allowing Cypress to provide superior fidelity to the original recording. With superior resolution in the time domain, Cypress can more accurately reproduce today’s advanced sound formats, including high-resolution and spatial audio.
Compared with legacy coil architectures MEMS parts show superior mechanical response, reduced phase shift. superior part-to-part phase consistency. The rigid silicon diaphragms eliminate speaker breakup for mid- and treble clarity. They are non-magnetic for lower weight and reduced electromagnetic interference
Cypress comes in a package measuring 6.3mm by 6.5mm by 1.65mm and is controlled by a separate ASIC dubbed Alta. Cypress is 40x louder in low frequencies compared to xMEMS prior generation MEMS speakers Cowell.
Cypress is made for xMEMS by foundry TSMC with a second-source foundry available. Housholder declined to name the alternate manufacturer at this time.
Full-function Cypress prototype silicon is sampling to selected customers. Production-candidate samples of Cypress and the companion Alta controller/amplifier ASIC will sample in June 2024.
xMEMS will demonstrate Cypress by appointment at CES 2024 in Las Vegas, Nevada, January 9 to 12.
xMEMS is not the only company to be exploring sound-from-ultrasound (see Startup uses MEMS ultrasound to improve audio speaker).
Related links and articles:
News articles:
Startup uses MEMS ultrasound to improve audio speaker
MEMS microphone market showing steady growth
xMEMS brings speakers plus programmable vent to production
MEMS microspeaker for wearables weighs 56mg
xMEMS Lab announces silicon speaker in production
Startup launches piezoMEMS speaker for TWS earphones
MEMS startup is part of Indian entry into chipmaking