The output of the accelerometer is strange.

[rsasaki0109]

@takahashi-e6

Hello. We just bought the following from you.

ADIS16470搭載IMUボード/ ADIS16475-2搭載IMUボード https://techno-road.com/products/tr-imu1647x.html

When I look at the readme.md of your ros driver, the accelerometer value seems to be output in [m/s].

$ rostopic echo /imu/data_raw
・・・
angular_velocity:
  x: -0.0116995596098
  y: -0.00314657808936
  z: 0.000579557116093
angular_velocity_covariance: [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]
linear_acceleration:
  x: 0.302349234658
  y: -0.303755252655
  z: 9.87837325989
linear_acceleration_covariance: [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]
・・・

Also, your ros driver seems to be converting from [G] to [m/s^2].

accl_[i] = (double)num_data[i+3] * GRAVITY / acc_sensi_; // convert unit g to m/s^2

https://github.com/technoroad/TR_IMU_LIB/blob/77245838adbfd14a7c87b961dfe2558d0abe5ea9/src/adis_rcv_csv.cpp#L261

However, when we check the output of the accelerometer with the imu in the horizontal position, it outputs roughly 1 in the z direction. This looks like the unit is [G].

rqt_bag plot

I think there is probably a bug in the acceleration value conversion. How can I solve this problem?

Thank you.

[takahashi-e6]

Thank you for contacting us.

After investigating, I found a suspicious part on the source, so I'll fix it. However, since there is no actual item at hand, after the actual item arrives, I would like to fix it.

We apologize for the inconvenience, but please wait for a while. Thank you.

[OsamuSekino]

@takahashi-e6 Hello.

I had a TR-IMU16475-2 on hand, so I tried it out. https://github.com/technoroad/TR_IMU_LIB/blob/dev/src/adis_rcv_csv.cpp#L525 Commenting out lines 525 through 529 seems to work correctly. If it helps.

Thank you.

[takahashi-e6]

@OsamuSekino Thank you for the information.

We are also paying attention to that part. However, there is a balance with other models, so I would like to fix it after seeing the operation on IMU16475-2.

Thank you.