Skip to main content
B
Buse
Visitor II
May 12, 2023
Solved

LSM6DSO32 calibration

  • May 12, 2023
  • 3 replies
  • 2705 views

Hello,

I made a special card design that includes LSM6DSO32. I want to use this sensor in motion classification project. I want to ask a question about a subject that is said by everyone but I do not know. 

Everyone has mentioned that it is necessary to perform a calibration process to get the correct data from an IMU. 

However, the LSM6DSO32 sensor's user manual states that it comes calibrated from the factory. Do I need to calibrate this sensor extra?

If yes, how should I do this for accelometer and gyro?

I read an article about the need to subtract the gyro values from the new gyro values by averaging at the stationary position, is this a correct way to calibrate?

I do not know how to calibrate the accelometer. I need help on how to calibrate both gyro and accelometer values.

Without doing any calibration, I set, the gyro is 1000 dps and accel is 8G and the sampling frequency of both to 104 Hz. The values I measured under these conditions are as in the picture. The IMU sensor's position is constant and perpendicular to the Z axis.


_legacyfs_online_stmicro_images_0693W00000bjwptQAA.pngThank for helping.

King regards,

Buse.

    This topic has been closed for replies.
    Best answer by Federica Bossi

    Hi @Buse​ ,

    I confirm you that the sensors are already calibrated by factory.

    However, for some applications, it could be suggested to subtract residual offset (for accelerometer) and ZRL (for gyro). To do so, acquire some samples for each axis and calculate the average. Then substract this value to the acquisitions.

    If this helps you, please mark my answer as "Best Answer" by clicking on the "Select as Best" button, this can be helpful for Community users to find this solution faster. 

    3 replies

    F
    Federica BossiAnswer
    Technical Moderator
    May 17, 2023

    Hi @Buse​ ,

    I confirm you that the sensors are already calibrated by factory.

    However, for some applications, it could be suggested to subtract residual offset (for accelerometer) and ZRL (for gyro). To do so, acquire some samples for each axis and calculate the average. Then substract this value to the acquisitions.

    If this helps you, please mark my answer as "Best Answer" by clicking on the "Select as Best" button, this can be helpful for Community users to find this solution faster. 

    B
    BuseAuthor
    Visitor II
    May 17, 2023

    Hi Federica,

    I'm asking you to confirm me, is it enough for me to take the average for each axis and subtract from the new value for both the gyro and the accelerometer? For the gyro this is true but for the accelerometer I'm asking because I'm not sure. Is it enough for me to calibrate by averaging in both (gyro and accelerometer) and subtracting from the new value?

    Could you explain a little more what you mean by some applications? I will use this IMU for motion estimation, is it necessary for this application? do i have to?

    Best regards,

    Buse

    F
    Federica Bossi
    Technical Moderator
    May 17, 2023

    Ciao Buse,

    I'm asking you to confirm me, is it enough for me to take the average for each axis and subtract from the new value for both the gyro and the accelerometer? For the gyro this is true but for the accelerometer I'm asking because I'm not sure. Is it enough for me to calibrate by averaging in both (gyro and accelerometer) and subtracting from the new value? Yes, it is.

    Could you explain a little more what you mean by some applications? I will use this IMU for motion estimation, is it necessary for this application? do i have to? It depends on how you implemented the algorithm. You can try not doing the calibration and in case add it later.

    B
    BuseAuthor
    Visitor II
    May 18, 2023

    Hi Federica,

    Thank for helps.

    Buse.

    Terms & ConditionsAccessibility statement