on - edited on by
Lina DABASINSKAITE
Summary
This knowledge article explains how the anti-alias filter ensures LIS2DU12, LIS2DUX12, LIS2DUXS12 data quality at low power consumption. A typical use case of an anti-alias filter is tap/double-tap recognition in wearable devices applications.
Introduction
Aliasing is a common source of error, which occurs when high frequencies fold in the band due to low sampling frequency. The solution could be working at high sampling frequencies, but this would increase the power consumption. So, aliasing can compromise the data or require additional post processing, increasing overall power consumption.
How to avoid aliasing in applications such as tap/double-tap recognition in wearable devices?
Always on the anti-alias filter (AAF)
The solution is an aggressive always-on anti-alias filter in the analog domain embedded in the smart accelerometers LIS2DU12, LIS2DUX12, and LIS2DUXS12, which ensures data quality at low power consumption.
A typical use case is the tap/double-tap recognition in wearable devices. On this purpose, the following picture shows multiple tap events collected at different ODR. A tap is recognized each time a positive peak is followed by a negative one.
The first row shows the ODR, which has been lowered down to 416Hz with no significant information loss. 400Hz is the suggested ODR for tap and double tap do be correctly detected.
The second row shows that whilst the AAF is still active, the ODR has been further reduced. In such configuration, the signal dynamic is loss and detection thresholds should be lowered increasing the risk of false detection until the algorithm is unworkable.
In the third row, the AAF is disabled and the consequent in-band folding compromises the information in the data stream that becomes unreliable (some negative peaks disappear).
Conclusion
This always-on anti-alias filter can avoid aliasing affecting tap/double-tap recognition in wearable devices at low power consumption.
Stay one step ahead with smart accelerometers LIS2DU12, LIS2DUX12, and LIS2DUXS12.
You can find a dedicated webinar on ST smart accelerometers at the following link: Webinar: An intelligent sensor for sustainable always-aware applications
For further details on the implementation of the AAF inside ST smart accelerometers, you can refer to:
- Paragraph 3.1 of the application note AN5648 for LIS2DU12;
- Paragraph 3.2 of the application note AN5909 for LIS2DUX12;
- Paragraph 3.2 of the application note AN5812 for LIS2DUXS12.