Following a very successful presence at Sensors Converge 2024 in Santa Clara, CA (June 25-26), where TDK promoted its range of MEMS microphones and sensors, the company announced worldwide distribution of its InvenSense SmartSound T5848 bottom port I²S MEMS microphone. This device is ideal to enable intelligent keyword, voice command and sound detection applications in ultra-low power designs, such has hearables and wearables. Originally announced at CES 2022, this MEMS microphone only now is available for general distribution.
The TDK InvenSense SmartSound T5848 is a low noise, high AOP I²S microphone with a 24‑bit PCM output for direct connection to a wide variety of SoCs. Supporting multiple operational modes, such as Sleep, Low Power, and High Quality, it also features Acoustic Activity Detect (AAD), and a new ultra-low power edge processing feature consuming as low as 20µA. This means that this is a MEMS microphone that is ideal for hearables and connected devices that will feature edge AI applications including wearables, TWS earbuds, glasses, smart speakers, and many other systems that are now featuring AI-powered features.
When the T5848 MEMS microphone was originally announced in 2022, together with several other microphones, voice interface applications had bottomed out, and reality had settled in about the lack of viable business models for the first, heavily cloud-dependent voice assistants. Fast-forward to 2024, and developers are now designing voice features on even lower-power designs that are based on Edge-AI platforms and make reliable voice interactions a possibility.
The T5848 MEMS microphone is a great device in those development scenarios and that’s likely the reason why TDK now announced that it would make it available through worldwide distribution channels. Together with the existing InvenSense SmartSound T5838, the T5848 I²S microphone can now support edge and generative AI systems with its key features and advantages, including low-power Acoustic Activity Detect (AAD), which is ideal for smart TV remotes and TWS earbuds. AAD allows these microphones to be programmed to listen for particular sounds, such as doorbells, keywords or voice commands, allowing the application’s processor to remain idle, and edge AI devices remain in an always-on state, ready to respond to user interactions without draining the battery.
In high-quality mode, the T5848 and T5838 MEMS microphones deliver 68dBA SNR and 133dB AOP, providing excellent abilities to ensure accurate keyword detection even in noisy environments while only consuming a minuscule 330µA at 1.8V. They consume only 130µA in always-on low-power mode, prolonging the battery life of always-on systems. Unique to the T5848 is the I²S interface support, which saves components in the system. It also reduces processing requirements such as filtering the microphone output in the system hardware or software.
For low-power edge AI applications, the exceptional SNR of the T5848 ensures systems receive high-quality input, even in noisy surroundings where keywords or voice commands must be distinguished from irrelevant background noise. Such high-quality input improves the accuracy and reliability of response required for next-generation AI systems.
“Voice interface is now ubiquitous in IoT devices as progress in AI technology makes voice interactions more natural and user friendly,” says Uday Mudoi, VP & GM – Microphone Business Unit, InvenSense, a TDK Group Company. “The T5848 with I²S interface enables simpler designs for always-on edge and generative AI systems.”
TDK T5848 microphones are available now for evaluation and integration, with developers benefiting from multiple ecosystem partners, which reduces required engineering investment and time to market for edge AI systems. The TDK InvenSense SmartSound T5848 MEMS microphone is supported by the same SmartSound One development platform already available for the T5838/T5837 devices.
Sensors Converge 2024
TDK demonstrated its MEMS microphones and the unique AAD feature at the Sensors Converge 2024 event that took place at the Santa Clara Convention Center, June 25-26, together with a full range of sensor system solutions targeted for applications across automotive, consumer, health, industrial, machine learning, robotics and more.
TDK experts demonstrated solutions featuring TDK’s magnetic sensors, MEMS sensors, temperature and pressure sensors, ultrasonic sensors, and positioning software for a wide variety of applications, such as robotics, wireless power transfer, energy harvesting, machine learning (ML), Internet of Things (IoT) and automotive — including e-mobility, advanced driver assistance systems (ADAS), and positioning.
The company announced the availability of two high performance InvenSense motion sensors, including the new InvenSense SmartMotion ICM-42370-P, a premium 3-axis accelerometer for consumer applications that require low noise and high sensitivity, and the IIM-42653, a high-FSR 6-axis IMU for industrial applications.
The new SmartMotion ICM-42370-P accelerometer offers low noise performance and up to ±16g of measurement range. It features a 16-bit ADC for high accelleration sensitivity and supports a wide range of on-chip motion event/activity detection based on interrupt or register access, without additional user programming. These features make it ideal for use cases in consumer designs, such as TWS earbuds, and all wearables.
invensense.tdk.com
www.tdk.com
Source link
