Portable devices call for more efficient inertial sensors
While the image quality of mobile phone cameras continues to improve, manufacturers still face challenges with image stabilization. Many people use software to improve overall sharpness, but still can't achieve the effect of a physical image stabilizer. In a mobile phone camera scenario, the movement of the camera is sensed by inertial MEMS while the image sensor moves in the opposite direction to compensate. But the tight housing and power consumption constraints of smartphones can present challenges to the use of inertial MEMS. AR and VR headsets also require more efficient sensors. When tracking head or hand movements, accuracy is critical. Accurate and fast sensors provide a more realistic experience and can even alleviate VR motion sickness. However, since most of these devices are powered by batteries, the requirements for energy consumption are getting higher and higher.
Engineers encounter practical difficulties when designing energy-efficient inertial sensors. To improve performance, engineers try to use filters and other mechanisms to reduce the signal-to-noise ratio, but this also increases power consumption. Engineers need to make trade-offs between stability and battery life. In addition, because MEMS must be packaged in small packages, engineers cannot increase the accuracy of MEMS by increasing the size.
Low power mode consumes only 0.65 mA
The LSM6DSV16X solves this problem with a new spring design in its mechanical structure. In addition, STMicroelectronics has adjusted the gain of the amplifier to increase performance while maintaining low power consumption. Therefore, the LSM6DSV16X consumes only 0.65 mA in high-performance mode (gyro and accelerometer power consumption combined), compared to 1.2 mA for the LSM6DSRX, although both devices have similar noise levels in low-power mode . Engineers developing optical image stabilizers for smartphones no longer have to contend with performance versus power consumption. By contrast, competing devices consume at least twice as much power in low-power modes as STMicroelectronics.
MLC runs 2x faster and supports Adaptive Configuration (ASC)
The Machine Learning Core (MLC) is another feature that helps save power. Running the sensor information with a decision tree eliminates the need to wake up the microcontroller, resulting in a significant reduction in overall power consumption. The MLC in the LSM6DSV16X operates twice as fast as its predecessor, and the output data rate (ODR) is increased to 100 Hz to 240 Hz. As a result, engineers can process more data and save more energy in a given time.
Users can also use 16 finite state machines (FSM) to identify specific patterns. Additionally, this new device supports adaptive configuration, enabling the FSM to reconfigure itself without waking the MCU. Developers can program various conditions and environments that will trigger reconfiguration for greater flexibility. Finally, both the MLC and the FSM can notify each other, enabling smarter applications and more accurate tracking of physical activity, such as detecting whether a smartphone is placed face down.
Qvar opens doors for new applications like people counting
The LSM6DSV16X brings Qvar to the LSM6 family of MEMS inertial sensors. Engineers can measure changes in ambient electrostatic charge by simply connecting two electrodes and enable this feature by changing two registers. Qvar opens the door to new applications such as people counting. Engineers can place electrodes on walls to measure proximity without using LEDs and photodiodes. We have also written corresponding application notes to help designers develop demo models. In addition, ST will provide more use cases by the end of the year. The LSM6DSV16X is pin-compatible with other members of the LSM6DS family through a pin-sharing mechanism between the Qvar and test pins.
LPS22DF and LPS28DFW
The water sensing problem
With the widespread adoption of wearable devices, activity tracking has become standard, and pressure sensors must be adapted to a wider range of physical scenarios. From mountain climbing to swimming, sensors must be able to reliably track users, providing accurate information on their physical activity. The challenge here is that climbing mountains or stairs is completely different than swimming in a pool, because the water creates different pressures. For example, diving to a depth of 10 meters in water exerts a pressure of 14.5 PSI, the same as atmospheric pressure at sea level. Therefore, pressure sensors must account for vastly different physical conditions without incurring excessive power consumption.
Consumes only 9.1 µA and 7 µA in high performance mode
The LPS22DF and LPS28DFW, the first pressure sensors to support dual full scale, help solve this problem. They can detect whether the user is above or below the water and switch to the appropriate range. The LPS22DF supports pressures up to 1,260 hPa, while the LPS28DFW supports pressures up to 4000 hPa, consuming 9.1 µA and 7 µA respectively in high performance mode. In contrast, LPS22HB consumes 12 µA at a maximum pressure of 1260 hPa.
LIS2DU12 LIS2DU12
The LIS2DU12 achieves a new balance between performance and power consumption. This product is more accurate than its predecessor thanks to the new LC filter and anti-aliasing filter. The LC filter can filter electrical noise and the anti-aliasing filter can prevent sampling errors, both of which can greatly improve signal quality without significantly increasing power consumption. The previous generation LIS2DW12 consumes only 380 nA in low power mode, while the newer, more accurate LIS2DU12 consumes only 450 nA. By comparison, the power consumption of competing products tends to hover around 1 µA.
