Maxic launches MT3102, a miniature, low power proximity sensor integrating skin recognition

MT3102, launched by Maxic, is a miniature, ultra-low power digital optical proximity sensor integrating 940nm VCSEL. The sensor is available in 1.4mm*1mm*0.35mm optimized package size and consumes only 0.7µA current in standby mode. By adopting the self-developed high accuracy skin recognition algorithm, the chip achieves >98%@3mm in-ear recognition rate, and <5%@3mm false-recognition rate for non-skin tissues such as desks, books, clothes, metal and so on. MT3102 is mainly used in wireless Bluetooth earphones such as TWS/OWS, portable hardware products, smart watches, smartphones, etc.

• Dedicated miniature package for small wearable devices

• Standby power<0.7µA, operating power<10µA

• Designed with innovative skin recognition algorithm, recognition rate>98%@3mm

• Strong noise and ambient light suppression capabilities, SNR>20@extreme scenarios

• Integrates VCSEL with 5mA driving current

• Supports 2.5V VCC input

• Flexible setting of VCSEL pulse width and count 

• 10~16 bits selectable ADC resolution

• Supports 1.2V I/O
  

MT3102 typical application circuit

■  Dedicated miniature package: 1.4mm*1.0mm*0.35mm

By integrating infrared light receiver, laser transmitter, I²C interfaces, high accuracy ADC and programmable low noise amplifier into an OLGA6 package with dimensions of 1.4mm*1.0mm*0.35mm, MT3102 is the smallest sensor in the industry, only accounting for 16% of the previous generation product’s volume.

MT3102 has the advantages of design flexibility, high performance and low power dissipation, and it is very suitable for wearable devices such as TWS earphones due to its ultra-simple peripheral circuit and compact design.

■ Ultra-low power: <0.7uA@standby mode

MT3102 is engineered with industry-leading power specifications, including operating/standby/sleep currents, where standby/sleep power<0.7µA, operating current<10µA@100ms (including VCSEL). The power has significantly decreased compared with the previous generation product, MT3101.

By adopting self-developed skin recognition algorithm, MT3102 is able to recognize more than 98% of the human body within 3mm, or more than 95% of other non-skin tissues such as clothes (white/black), paper (in different colors), tables (wood/metal) or other electronics (with plastic shell), greatly enhancing the wearing detection performance and improving the user experience.

MT3102 is designed with industry-leading SNR, minimum package size and excellent crosstalk suppression capabilities, including <100LSB@16bit package crosstalk and <300LSB@16bit for 0.2mm under glass; integrates OTP and supports factory calibration to eliminate DC error after assembly. Additionally, the chip incorporates 940nm VCSEL widening the Field of View (FOV) to capture object signal changes over a broader range without impacting Crosstalk (CT). This design enables wider proximity sensing capabilities.