The World’s First Truly Monolithic MEMS Speaker Hits the Market

July 17, 2020 by Jake Hertz

MEMs speakers already have a leg up on conventional speakers. What does a fully-monolithic MEMs speaker entail?

By most standards, conventional speaker architectures are clunky. They rely on many large components like coils and magnets to convert electrical energy into mechanical energy—compressing air and producing audio. 


Parts in a conventional loudspeaker.

Parts in a conventional loudspeaker. Image used courtesy of Headstock Distribution

In a world of constantly miniaturizing technology, engineers are looking to MEMS speakers as a viable alternative.


The Genius of the MEMs Speaker Design

MEMs technology has been around for the past three decades for devices such as microphones, accelerometers, vibration sensors, and many others.

Only recently has it been applied to speaker design; the results is significantly smaller and lower-power speaker systems. 


Audio Pixels MEMS speaker technology.

Audio Pixels' MEMS speaker technology. Image used courtesy of Mike Klasco


MEMS speakers work by using piezoelectric thin films as active elements that deform when a voltage is applied. This mechanical deflection displaces the surrounding air and thus generates sound waves.

It turns out that this technology is well suited for silicon, so we have the means to produce it cost-efficiently and in very small areas. 


Benefits of MEMs Speakers Over Conventional Speakers

MEMS speakers offer a slew of benefits over conventional speakers.

For starters, their small size and manufacturing process makes them easily integrated into PCBs. By the same token, these speakers can be placed on the same PCB substrate as their audio amplifier. This makes for extremely size-efficient designs that are essential for wearables like earbuds. 


USounds integrated an amplifier and MEMS speaker on the same PCB substrate.

USounds integrated an amplifier and MEMS speaker on the same PCB substrate. Image used courtesy of Digi-Key


On top of this, MEMS speakers are considerably lower power thanks to their high intrinsic impedance. In tests comparing MEMS to a reference balanced armature and electrodynamic speakers, MEMS consumed the least power of the three by a pretty good margin.


Test results comparing MEMS power consumption to a balanced armature and electrodynamic speakers

Test results comparing MEMS power consumption to a balanced armature and electrodynamic speakers. Image used courtesy of Digi-Key

This too will be important when designing for wearables since they operate with small batteries and thus need maximized power efficiency. 


The First Monolithic True MEMs Speaker?

Last week, xMEMS announced the release of Montara, which the company claims is the world’s first monolithic true MEMS speaker.

Up to this point, many MEMS speakers on chips have been hybrids, where the actuator and diaphragm were not implemented in silicon. 


xMEMs technology

xMEMs technology. Image used courtesy of xMEMs

This new speaker produces some pretty impressive specs. It offers a full bandwidth (20 Hz to 20 kHz) with a flat frequency response and in-band total harmonic distortion (THD) of less than .5%.


The new Montara MEMS chip

The new Montara MEMS chip. Image used courtesy of xMEMS


Along with this, xMEMs says Montara is the world's first IP-57-rated microspeaker, enabling water- and dust-resistant earbuds. 


The Future of Fully-Monolithic Integration

Montara represents a huge achievement, accomplishing a monolithic full-bandwidth, low-power, and easily-integrated MEMS speaker. These impressive specs along with its small size and waterproofing capabilities make Montara a notable presence on the market.

Dimitrios Damianos, Technology & Market Analyst at Yole Développement, says, "The replacement of traditional voice coil speakers and balanced armatures with a complete semiconductor component provides a scalable design that claims significantly lower power and latency, while xMEMS' fabless model ensures high volume repeatability.” 

The hope is that with this technology finally achieving fully-monolithic integration, we’ll start to see similar devices utilized in coming generations of wearable technology.