update to imu docs

[mbz4]

Expanded configuration to include manual workaround to IMU misaddressing & calibration bug.
Adjusted structure to improve overall information flow.

[Ezward]

Thank you for these changes; there are a lot of nice improvements. See my comment regarding changing the IMU address; I've made that much easier.

[Ezward]

I've actually landed a pull requests that exposes a configuration so there is not need to edit this manually; just change the configuration value; See autorope/donkeycar#1064. The configuraton changes here https://github.com/autorope/donkeycar/pull/1064/files#diff-dd7806493c6284e60142951760a6712d8b3f90a8e007e710d4033691f74753edR447. So it would be excellent if you could incorporate instructions for changing that configuration here.

[mbz4]

Tried to understand what's happening at the provided recent merge.
==> configurable imu addressing (?); derives from imu part file w/ new (?) IMU_ADDRESS config (?)

Thus, and please do correct me, @Ezward

1. the imu docs should instead reflect setting addr=0x68 to addr=IMU_ADDRESS on line 29
   ... in addition to commenting out the calibrateMPU6500() script
   ... & setting address_mpu_slave = 0x0c
   ... not sure where IMU_ADDRESS would be initially defined here

Or

2. are you suggested the imu docs should additionally be extended to contain info on how to modify:
   IMU_ADDRESS = 0x68 # if AD0 pin is pulled high them address is 0x69, otherwise it is 0x68
   ...which can be found at: donkeycar/templates/cfg_path_follow.py (?)

[mbz4]

I think option 2 seems more likely.
So, I should include instructions how to configure IMU_ADDRESS in:

1. donkeycar/templates/cfg_path_follow.py::414

&

3. donkeycar/templates/cfg_complete.py::447

Appreciate the feedback

[Ezward]

I am suggesting 2. We no longer need to edit any source code to change the address; so that should be removed. The user just needs to set the value of "IMU_ADDRESS" in their myconfig.py file; so that should added.

[mbz4]

Hey, I studied the IMU_ADDRESS commit further to try and understand why source code changes to imu.py are no longer necessary and I did not reach the same conclusion; I left the source code edits in case configurations differ among setups

[Ezward]

See https://github.com/autorope/donkeycar/blob/f8dc720596d04c3cb14ff0a34d771ef39baf1897/donkeycar/templates/complete.py#L263 https://github.com/autorope/donkeycar/blob/f8dc720596d04c3cb14ff0a34d771ef39baf1897/donkeycar/templates/complete.py#L805 The imu part has be updated to take in the necessary parameters and complete.py has been updated to use them.

I also see that the instruction include setting the address_mpu_slave=0x0c; can you explain? My understanding what that the slave setting is if you have two IMU's and one is slaved to the other.

I also see that the instructions include commenting out the sensor calibration; can you explain? The calibrateMPU6500 is still appropriate even if using an MPU9250; the MPU9250 is basically just an MPU6500 plus an magnetometer; we we still want to calibrate the giro/accelerometers so we know which way is down.

[Ezward]

See my comment; I think we still need to hash out the manual changes.

[Ezward]

See https://github.com/autorope/donkeycar/blob/f8dc720596d04c3cb14ff0a34d771ef39baf1897/donkeycar/templates/complete.py#L263 https://github.com/autorope/donkeycar/blob/f8dc720596d04c3cb14ff0a34d771ef39baf1897/donkeycar/templates/complete.py#L805 The imu part has be updated to take in the necessary parameters and complete.py has been updated to use them.

I also see that the instruction include setting the address_mpu_slave=0x0c; can you explain? My understanding what that the slave setting is if you have two IMU's and one is slaved to the other.

I also see that the instructions include commenting out the sensor calibration; can you explain? The calibrateMPU6500 is still appropriate even if using an MPU9250; the MPU9250 is basically just an MPU6500 plus an magnetometer; we we still want to calibrate the giro/accelerometers so we know which way is down.