Differences between acceleration measurements on three axis

By dan_leland , 6 February 2012

Hi,
with the EV_MPU6050_A, I tried to collect acceleration measurements for each axis (using a spirit level for best accuracy) in six different position (six face position test):


    1°: z axis upward -> max value: 7668
    2°: z axis downard -> min value: -8700
    3°: y axis upward -> max value: 8330
    4°: y axis downard -> min value: -8314
    5°: x axis upward -> max value: 8300
    6°: x axis downard -> min value: -8106


and then calculate the mean values as (axis_up+axis_down)/2 for each axis. But, while for x and y axis the values is close to 0, for the z axis I obtain: -516. In short, the module of acceleration on z axis is very different when I place z upward and z downard. Are these difference normal?

In addition I've made also a data acquisition of 15h @ 200 Hz (about 10 million of samples) with z axis upward:
Image removed.
Is this drift normal?

Regards
phpbb Topic ID
14115