I'm try to understand what can be the real usable resolution of the LIS3DSH accelerometer because I need to evaluate it for inclometric measurements.
From the datasheet I can read a PSD=150ug/sqrt(Hz) @ ODR=100 Hz then BW=50 Hz
For a analog output with one single pole filter the overall noise can be calculated as PSD*SQRT(1.57*BW),supposing to use same calculation we have 150ug*SQRT(78.5)=1.33mg, having a sensitivity equal to 0.06 mg/digit @ ±2g we have:
1.33mg/0.06 mg/digit = 22.16 digit of error then instead the theoretical 16 bit we should have 65536/22.16=2957.4 digit which is approx equivalent to 11-12 bit of final resolution.
I kindly ask if this is a correct mode to estimate the real final resolution by taking in account the effect of the noise or if I missing some point and then what can be the correct way to perform this calculation.
Hope someone can point me in the correct direction.
Thanks in advance and