2025-01-16 06:30 AM
I want to recognise small movements with small acceleration.
Which mode is best suited?
Is it better to have a high sampling rate or a lower one?
which filters?
where do the large peaks (incorrect values) come from?
But I can already recognise the acceleration and braking
2025-01-16 08:13 AM
Hi @Stefan3 ,
A smaller FS range (e.g., ±2g) will provide higher sensitivity and better resolution for detecting small accelerations.
A higher sampling rate can capture more data points and provide a more detailed representation of the movement. This is useful for detecting small and rapid changes in acceleration. However, it will also increase power consumption and the amount of data to process.
I suggest you to use High-accuracy ODR (HAODR) mode. This mode typically offers better precision and lower noise.
About the incorrect values, did you enable the BDU bit?
2025-01-17 06:09 AM
Hi Federica
±2g is set.
I haven't done anything yet
except set CTRL1(10h) to 960Hz and read with 10mSec (100Hz) with the I2C bus (1MHz).
I have enough power available but reading faster than 100Hz is not so easy.
Should I choose low-power mode3 (8mean)? Or a high performance mode?
Should I enable the BDU bit anyway?
2025-01-17 06:25 AM
Hi @Stefan3 ,
Yes, it is always suggested to enable the BDU bit.
In addition, high performance mode is suggested for this kind of application :)