By: Mike Housholder, xMEMS Labs, Inc.
What is solid-state fidelity? First, letโs start with a couple of definitions. โSolid-stateโ is a common descriptor used to refer to electronic components, devices and systems based entirely on semiconductor materials such as silicon. โFidelityโ, as it relates to this article, defines the degree of exactness and precision with which audio is reproduced.
To those who have not followed the recent advancements in speaker technology, you are reading this correctly, it is now possible to produce rich, precise audio using only silicon and semiconductor manufacturing processes.
Figure 1: xMEMS solid-state MEMS micro speaker
In the consumer electronics industry, there is a natural, almost gravitational pull toward solid-state technologies. They are easier to manufacture at scale and offer enhanced reliability, performance and eventually lower cost as they meet high volume demands. Some recent examples of solid-state transition are the migration away from hard-disk drives to solid-state drives (SSDs), the evolution to silicon microphones from electret condenser type, and inertial sensors where spring-mass systems are forged out of micromachined silicon using semiconductor processes. Now it is time for the speaker to make the same transition. The mechanical coil/magnet speaker architecture has existed, almost unchanged, for nearly 100 years and the time has come for solid-state disruption.
Advancing the Consumer Audio Experience
The paradigm for what is โgood enoughโ is about to change as dramatically for your ears as HD video changed what was good enough for your eyes.
In a prior article, we described the consumer transition to Hi-Res audio and how solid-state MEMS speakers play a role. However, from the perspective of the speaker, the Hi-Res Audio standard focuses on the use of ultrasonic harmonics (>20kHz) to blend with the audible spectrum (20Hz-20kHz) to enhance the character of acoustic instruments in music.
What is not addressed with Hi-Res Audio is how to enhance sound reproduction in the audible spectrum through improvements in speaker technology.
This is where solid-state fidelity comes in. A new transduction mechanism and the first significant advancement in speaker performance characteristics in decades that can lead to a new era of ultra-high-resolution audio: enhanced clarity and detail, accurate reproduction of the source material with improved presence and space. The solid-state speaker can match or exceed the resolution (bit depth and sampling frequency) of the source audio stream delivered by the codec.
The Defining Performance Characteristics of Solid-State Fidelity
Solid-state MEMS speakers are defined by key attributes built around a unique speaker diaphragm material and the design of a distributed drive mechanism that combine to produce several unique performance characteristics.
1. Improved Diaphragm Material Stiffness
It is common practice in the hi-end tower speaker market to utilize speaker diaphragm materials with a high Youngโs Modulus (or material stiffness factor) in the midrange and tweeter drivers[1]. This is done to mitigate the speaker break-up phenomenon where the diaphragm exhibits non-uniform excursion, which creates muddy sound. Break-up is a function of material stiffness and weight and better materials can help to eliminate break-up effects or push break-up to higher, less critical frequency ranges. Where the woofer and subwoofers commonly use paper and plastic, the midrange and tweeter drivers require more exotic materials to avoid break-up effects and to deliver the best quality sound.
In the consumer earbud market, it is most common to find various types of plastic diaphragms in traditional dynamic drive speakers, making them susceptible to break-up.
Figure 2: Material stiffness chart of silicon vs. typical speaker diaphragm materials
xMEMS is the first to bring silicon diaphragms, a material with a premium stiffness factor (95x better than plastic), to consumer earbuds as a part of our solid-state process. This premium material yields impeccably accurate and clear midrange and treble presence for ultra-high-resolution audio.
2. Distributed Piezo Drive
Another key factor to mitigating speaker break-up is to improve how the speaker diaphragm is driven.
Traditional dynamic drive speakers are driven by a coil and magnet system, where a thin copper coil pushes on only a narrow section of the diaphragm to move air and generate sound. This is a โthin annulus concentrated forceโ drive, where most of the diaphragm is not supported directly by the drive mechanism, increasing the likelihood of non-uniform behavior and speaker break-up.
Figure 3: Dynamic drive speaker coil attached to plastic speaker diaphragm
xMEMS improves on legacy speaker architectures by replacing the concentrated coil/magnet drive with a monolithic piezo drive, where the piezo is implemented as a layer in the semiconductor process directly under the diaphragm. The piezo layer is distributed across the majority of the speaker diaphragm surface. This โdistributed supportโ reduces the chances of speaker break-up due to the diaphragm surface being driven uniformly by the underlying piezo materials.
This distributed drive also enables speaker designs with multiple resonances which helps to achieve wider acoustic bandwidth. By comparison, coil-based speaker operation is designed for a single mechanical resonance and results in limited bandwidth due to sound pressure level decay at frequencies higher than the resonance.
Figure 4: xMEMS piezo drive distributed across the speaker diaphragm surface
3. Fast Mechanical Response
When measuring the performance of a speaker, most of the focus is on amplitude (or sound pressure level) across frequency. But time is another key metric. When a speaker responds to an impulse, such as a drum rim shot, it should start instantly and stop the instant the instrument stops making sound. If the speaker keeps vibrating or resonating and making sound after the source sound stops, itโs changing, or โcoloringโ the sound of the original recording, which is undesirable.[2]
To realize the full potential of ultra-high-resolution audio, a faster and more precise transduction mechanism is required to deliver accurate sound reproduction that is true to the original recording.
Solid-state speakers are voltage-driven, not current-driven like traditional coil speakers. When voltage is applied to the piezo drive, it generates an instant impulse response with a very fast settling time, which avoids coloring and enables xMEMS speakers to deliver pulse-true reproduction.
Figure 5: Calculated impulse response of xMEMS speaker vs. typical TWS dynamic drive speaker; derived from the measured stimulus response on a Klippel Analyzer 3 (KA3) system.
4. Near-Zero Phase Shift
Random phase response shifts from the speaker can also distort the sound, especially with speech or the attack at the beginning and drop at the end of a sustained note. With dynamic drive coil speakers, a phase shift of 180ยฐ is common at or near its resonant frequency which is typically in the sensitive audio frequencies (500Hz-2kHz).
xMEMSโ solid-state speakers place the resonant frequency well outside the critical audio frequencies (>10kHz). This enables a flat, near-zero (2ยฐ) phase response that eliminates any non-linear effects in the sound.
Figure 6: xMEMS Montara 2ยฐ phase shift vs. 10mm TWS Dynamic Driver (DD)
5. Phase Consistency
Another benefit comes from the uniformity and consistency of the semiconductor manufacturing process. Each xMEMS speaker exhibits a part-to-part phase consistency of +/-1ยฐ. This is valuable to minimize matching of left and right speakers with minimal calibration. Moreover, for spatial audio applications, this consistency improves accuracy and separation and maintains audio clarity as audio moves from left speaker to right speaker and vice-versa. No speaker-to-speaker variation = no muddy response and more precise spatial resolution.
What Does Solid-State Fidelity Sound Like?
Ultimately, hearing is believing. At a recent CanJam event, the worldโs premier headphone and earphone audio show for audio enthusiasts, we assembled 3D-printed earbuds each with a single, full-range xMEMS speaker. These earbud prototypes were evaluated in listening tests by hundreds of show attendees. These are a sampling of actual comments from listeners at the July 30-31, 2022 event in London, UK:
Summary
Solid-state fidelity is not just a new way to build speakers, itโs something much moreโฆItโs an opportunity to raise the expectations for audio clarity, precision and reproduction accuracy for consumer and professional earbuds. It is time to move on from century-old speaker technology.
Solid-state MEMS speakers usher in the era of ultra-high-fidelity audio. Consumers should no longer accept the status-quo.
About xMEMS
xMEMS is reinventing sound with the worldโs first solid-state MEMS speakers for TWS and other personal audio devices. The monolithic architecture implements both actuation and diaphragm in silicon resulting in unmatched part-to-part frequency response consistency and reduced speaker matching or calibration time at manufacturing. This innovative transduction mechanism has also produced the worldsโ fastest and most precise ยตspeakers, eliminating spring and suspension recovery of coil speakers which improves audio quality and sound field reproduction. xMEMS speakers are mass-production ready. Please contact us to learn more. http://www.xmems.com; marketing@xmems.com
[1] Bowers & Wilkins: The Continuum Cone; https://youtu.be/Odru682HebY
[2] https://www.ecoustics.com/articles/understanding-speaker-frequency-response/