Tag Archives: LSM9DS0

ROS Hydro/Indigo and sparkfun IMU LSM9DS0 9DOF

I’ve connected the LSM9DS0 9 degrees of Freedom Breakout Board made by sparkfun with an arduino micro like I’ve described in a previous post, wrote a little rosserial sensor_msgs::Imu publisher and visualized everything using the rqt plugin manager for further experimenting.

Here a screenshot while moving the setup:

ScreenshotIMURos

a screenshot while not touching:

ScreenshotIMURos2

As you see all data is still moving a lot. So the next step beside finding out what the units really mean, will be stabilizing by using a kalman filter like its provided in the robot_pose_ekf package.

By the way: the arduino is currently using around 25000 of its 28 672 bytes memory just providing the IMU data to ros. So the motorshield will require another micro or we switch it to something else like an wiimote.

Arduino Micro and 3.3V IMU LSM9DS0 9DOF

Soldering, Soldering, Soldering 🙂 Everything else had been following the amazingly well written guides of the LSM9DS0 made by sparkfun. Nine degrees of freedom at a rate of “a few per second”(currently 9Hz) since I’ve followed just the basic setup without fancy interrupt usage.

One thing thats really important to mention is the different signal voltage level of the SDA and SCL pins between the Micro(5V) and the IMU Breakout Board(3.3V) – which in case you connect them together without bi-directional level shifting, as you might expect since i2c is designed for exactly that, would lead to blue chip burn.

So wiring on the bread board (and not removing the wires used by the arduino motor shield v2, so do not get too confused by that):

IMG_20140612_192126

and applying the library to the arduino IDE, leads to a working live example with 2 outputs per second:ScreenshotIMULSM9DS0So the next step is to increase the rate by improving the setup wiring, parse that data into ROS Hydro by a SensorMsg/Imu publisher,  kalman and combining these with other odom sources like my currently used (and sadly poor)  or even an GPS source to a exact and really usable Odometry by the robot_pose_ekf package for later Simultaneous Localization and Mapping (SLAM) – a real autonomous mapping and navigation. Sounds easy right?