Post Go back to editing

10-2000 Hz and 1,3 and 5 grms vibration tests cause bias shift in accelerometer channels.

Category: Hardware
Product Number: ADIS16475

We have run a test scenario for our new design product which includes ADIS16475-2. When we observe test results there is a problem shown in figures clearly.

1) TEST INFORMATION
IMU: ADIS16475-2
Firmware Rev : 0x0134
Month Day : 0x0325
Year : 0x2023
Serial : 0x1369
IMU Sample Rate: 1000 Hz
2) VIBRATION PROFILE
IMU is placed on circuitboard and has metal enclosure. The complete enclosure is placed on
vibration table. Frequency Range: 10-2000 Hz and 1,3 and 5 grms vibration tests are realised
without a mechanical isolator. The test table is vibrated in Z axis. IMU body axes and test table’s
axes are aligned.
3) TEST DATA
It is observed that, after the vibration occured, accelerometer data (all 3 axes) has considerable
errornous bias shifts. Only the Z axis acceleration data is shared below however the other axes
have the same characteristics. After each vibration test, it is expected that accelerometer
measures zero acceleration in X and Y axes and 1 g in Z axis.



Hi 

Thank you for sharing these plots.  I will get back to you after conducting some similar tests and we will go from there.

  • Hi,


    I want to share the test results for the same part that I tested in the lab for 1grms, 3grms, and 8grms. As you can see in the attached figure, I do not see any bias shift. I collected the data using ODR of 1000hz.

    I recommend ensuring the IMU is securely mounted and Identifying the resonant frequencies of the mounting setup (fixture). As mentioned in the datasheet, you are sweeping up to 2000hz, which is very close to the accelerometer's resonant frequency of 2.2khz.