Using BMM270 Accel/Gyro and BMI150 Magnetomer Sensors with Teensy 4.x
- Thread starter mjs513
- Start date
mjs513
Senior Member+
Was wondering about that when I saw it in firmware for the m5stack - didn't see it on the schematic though.
mjs513
Senior Member+
@russellkt
Received my Atoms3r and been playing with it. Finally got it figure out. But for some documentation of whats going on.
1. Took apart the case to determine chip orientation instead of guessing.
2. Looking at the datasheets for pin1 /axis orientation and putting in NED Coordinates gives the following conversion
3. But to do item 2 you have to edit the imu_class.cpp file in m5unified/src/utility directory
change it (about line 89) to
Notice I commented what was there just to keep track. Probably easier way to do but at least it works for now.
Then after updating the sketches I put together (will make them attachments)
Motion gives
For acceleration I used 6point cal: which gives
And fusion then becomes stable and compass is now working.
Files so you can play are now attached
Cheers Mike
Received my Atoms3r and been playing with it. Finally got it figure out. But for some documentation of whats going on.
1. Took apart the case to determine chip orientation instead of guessing.
2. Looking at the datasheets for pin1 /axis orientation and putting in NED Coordinates gives the following conversion
C++:
// Match axis orientation from example
val[0] = -ay;
val[1] = -ax;
val[2] = -az;
val[3] = -gy ;
val[4] = -gx ;
val[5] = -gz ;
val[6] = mx / 10.0f;
val[7] = my / 10.0f;
val[8] = mz / 10.0f;3. But to do item 2 you have to edit the imu_class.cpp file in m5unified/src/utility directory
change it (about line 89) to
C++:
#if defined ( CONFIG_IDF_TARGET_ESP32S3 )
if (board == m5::board_t::board_M5StackCoreS3 && bmi2->getAddress() == 0x69)
{ // CoreS3 では、地磁気のY軸Z軸をそれぞれ反転する
_internal_axisorder_fixed[sensor_index_mag] = (internal_axisorder_t)(axis_invert_y | axis_invert_z); // Y軸,Z軸反転
} else
if (board == m5::board_t::board_M5AtomS3R || board == m5::board_t::board_M5AtomS3RCam || board == m5::board_t::board_M5AtomS3RExt)
{ // AtomS3Rシリーズ では、ジャイロと加速度のY軸とX軸を入れ替え、Y軸を反転するほか、地磁気のX軸とZ軸をそれぞれ反転する
//Then, the gyro and acceleration Y-axis and X-axis are swapped, and the Y-axis is reversed, and the X-axis and Z-axis of the geomagnetic field are reversed, respectively
// _internal_axisorder_fixed[sensor_index_accel] = (internal_axisorder_t)(axis_order_yxz | axis_invert_y);
// _internal_axisorder_fixed[sensor_index_gyro ] = (internal_axisorder_t)(axis_order_yxz | axis_invert_y);
// _internal_axisorder_fixed[sensor_index_mag ] = (internal_axisorder_t)(axis_invert_x | axis_invert_z); // X-axis, Z-axis inversion
_internal_axisorder_fixed[sensor_index_accel] = (internal_axisorder_t)(axis_order_xyz);
_internal_axisorder_fixed[sensor_index_gyro ] = (internal_axisorder_t)(axis_order_xyz);
_internal_axisorder_fixed[sensor_index_mag ] = (internal_axisorder_t)(axis_order_xyz ); // X-axis,
}
#endifThen after updating the sketches I put together (will make them attachments)
Motion gives
For acceleration I used 6point cal: which gives
Code:
Accel X Bias: 0.046889
Accel X Scale Factor: 1.001222
Accel Y Bias: 0.046889
Accel Y Scale Factor: 1.003922
Accel Z Bias: 0.040066
Accel Z Scale Factor: 0.997322And fusion then becomes stable and compass is now working.
Files so you can play are now attached
Cheers Mike
mjs513
Senior Member+
Nope. Forgot o post it: https://x-io.co.uk/x-imu3/ scroll down to the downloads section
I did and got the following in MotionCal:
All looks good. Also went with the six point calibration and got:
I did not use the magnetometer scale factor/bias.
However, when I show in imu3 all looks ok but after a yaw movement, it will reach a position, and then drift away when holding steady. Created the attached sketch and heading is all over the place.
All looks good. Also went with the six point calibration and got:
I did not use the magnetometer scale factor/bias.
However, when I show in imu3 all looks ok but after a yaw movement, it will reach a position, and then drift away when holding steady. Created the attached sketch and heading is all over the place.
mjs513
Senior Member+
Strange - mine was steady.
PS looked at your sketch and did see where you included the accel caibration. You also didn't include a few things that you need. Will give it try a bit later - have a few things to do outside programming right now.
mjs513
Senior Member+
@russellkt
Spent some more time on this and got my library working with the AtomS3R. All files for using the Library with the S3R and the files for working with the builtin lib have been posted to:
github.com
Spent some more time on this and got my library working with the AtomS3R. All files for using the Library with the S3R and the files for working with the builtin lib have been posted to:
GitHub - mjs513/M5-AtomS3R-IMU
Contribute to mjs513/M5-AtomS3R-IMU development by creating an account on GitHub.
github.com
Cannot express how thankful I am for all of your work. Tied up with kids at Christmas today but hopefully can get some time tonight or tomorrow.@russellkt
Spent some more time on this and got my library working with the AtomS3R. All files for using the Library with the S3R and the files for working with the builtin lib have been posted to:
GitHub - mjs513/M5-AtomS3R-IMU
Contribute to mjs513/M5-AtomS3R-IMU development by creating an account on GitHub.
Thank you again and Happy Holidays!!
Sorry for the delay. Thank you again for all of your work! Unfortunately, I adjusted the orientation in the m5 library, calibrated the atom using motioncal, inputted the values, and am still getting very sketchy results. Compass only goes between ~150 and ~190 degrees. At this time, I'm I will probably abandon for the time being until I can get another one to rule out any board or sensor issues.
mjs513
Senior Member+
Thats strange seems to be working for me. Did you use the updated sketches I had for MotionCal?