Understanding Electrical Line Resistance at Advanced Semiconductor Nodes
November 16, 2021
The Effect of Pattern Loading on BEOL Yield and Reliability during Chemical Mechanical Planarization
December 3, 2021
In a previous blog post published in September 2020, I stated that MEMS microphone demand is being driven by customer-requested audio features such as stereo sound, voice recognition, sound directivity, noise cancellation and more. To implement these audio features, multiple MEMS microphones are needed to support this functionality within the finished product. For example, a state-of-the art smartphone might include up to 6 microphones. The demand for MEMS microphones has accelerated in part due to a recent increase in remote work activities. Remote office work and online meetings are becoming the new norm, and these meetings would not be possible without good voice input systems (i.e. – microphones).
Voice sensors are also being used in True Wireless Stereo (TWS) earbuds (Figure 1). These earbuds can wirelessly connect to each other and to a smart phone or other Bluetooth-enabled appliance. Each stereo earbud contains a speaker (which in the future may be a MEMS micro-speaker) but also multiple MEMS microphones. Most high-tier TWS systems include three MEMS microphones per earbud: two external microphones that are used for beam forming and one internal microphone used for Active Noise Cancellation (ANC). The two external microphones must be slightly distant from each other, which explains why a long bar is included in the typical P-shaped TWS earbuds (Figure 1).
Figure 1: Commercial True Wireless Stereo earbuds (Courtesy: Apple, Inc.).
Active Noise Cancellation (ANC), a desired feature in wireless earbuds, successfully reduces background and wind noises but also reduces sound reception. Due to this problem, ANC is often turned off during voice calls [1]. To overcome this limitation, a new type of MEMS device is being used to directly capture voice through bone and skull vibrations. These new devices are called Voice Pick-Up Sensors (VPU) [2] or Voice Pick-Up Accelerometers [1] or simply Accelerometers [3]. They are comprised of a MEMS accelerometer that has the frequency characteristics of a MEMS microphone. Basically, the device is a hybrid mix of a microphone and an accelerometer. The VPU sensor is comprised (in a simplified form) of a modified MEMS microphone with a mass attached to the microphone membrane and a sealed sound-port (Figure 2).
Figure 2: Schematic view of MEMS microphone (left) and Voice Pick-Up (VPU) bone sensor (right). Note the added mass and absence of a sound port on the VPU.
Along with improved voice detection and background noise immunity, VPUs have a number of additional advantages in Active Noise Cancellation applications:
- Tapping detection is supported. This offers an interactive way to control a device (e.g. tap it once to increase speaker volume, tap it twice to decrease volume),
- Barge-in capabilities. The VPU can pick-up voice activity even when the internal speaker is playing music. In other words, conversation is detectable even while listening to loud music.
- Since no sound port is used or required, the VPU can be fully sealed in its package and is better protected from external environmental contamination.
Figure 3 MEMS+ VPU Model (a quarter of the device is shown)
MEMS+®, a tool of the CoventorMP® MEMS design platform, can be used to design MEMS microphones, accelerometers, and Voice Pick-Up (VPU) sensors. A quarter of a MEMS+ VPU model is shown in Figure 3. This model can accurately reproduce nonlinear device behavior, which is critical during circuit and system level design. The model is also fully parameterized and supports multi-physics modeling. Recently, we released CoventorMP 2.0, the latest version of our MEMS design software. The new software allows the use of charge sources in your MEMS design, which is key to testing MEMS microphones and VPUs in real life configurations. Each of these capabilities in CoventorMP can be used in the design and development of highly advanced, next generation voice detection devices.
The demand for MEMS microphones will continue to increase, due to the introduction of new applications and new products incorporating voice sensing. New MEMS-based voice sensors (such as VPUs) that support advanced functionality, along with additional demand for new and better voice sensing solutions, will continue to fuel the growth of next generation MEMS microphones and MEMS-based voice sensing products.
References:
[1] Vesper white paper, Voice Accelerometers in TWS Earbuds, A Sensor and Algorithm Perspective
[2] Paul Clemens, Sonion Voice Pick up (VPU) Sensor, TDK Developers Conference 2018, Santa Clara Marriott, September 17-18, 2018
[3] ST-Microelectronic brochure, LIS25BA, Low-noise, high-bandwidth accelerometer with TDM interface
![Fig. 2. Silicon spin qubit centered in the dotted circle, control and readout signals (M, P, R, T, and Q) are shown in the inset. Simplified schematics of the quantum point contact and corollary circuits are shown. The voltage source is implemented as a digital-to-analog converter at room temperature. (Taken from [6])](https://www.coventor.com/wp-content/uploads/2022/05/Fig.-2.-Silicon-spin-qubit-centered-in-the-dotted-circle-control-and-readout.png)
![Figure 2: Schematic view of standard CMOS including an electro-mechanical switch monolithically integrated into BEOL [2]. Courtesy: University of California, Berkeley](https://www.coventor.com/wp-content/uploads/2022/04/Figure-2-April-2022-MEMS-Blog-for-Coventor-website.png)
