So for the past week I have been lurking through the forums searching for people with the same problem and have found quite a few that have post here previously. I have tried to look at the input and check to see where I went wrong by the suggestions and yet still a bit confused.

When reading the magnetometer, from my understanding the number you will read from it will be between -4027 and 4028 with those being overflow conditions. I seem to be receiving huge numbers from the magnetometer (etc X MAX: 32512 MIN: -32512). But from the datasheet I wouldn't think this would be a problem seeing as it is stated the magnetometer data is 16-bits.

I thought ok I will set the gain down on the compass and got readings like 2034 and -2141 which I thought maybe that would be better. The problem I have is still present. Yaw wants to drift to N which i turn off drift detection my Yaw is extremely solid. Pitch and Roll seem to always be very solid except when Roll reaches 90 Pitch is shoot up to 90 also.

If anyone has any suggestions please let me know. The code I have is the basic Pololu MinIMU code for the arduino.

I found out that I had the High byte and Low byte swapped on the compass. The high byte comes first as opposed to the accelerometer which the low bytes is first. I re-calibrated the compass and got values that are looking like normal. My problem is still when I move the sensor around Yaw it will always move back to North.

Hi Brion, have you solved your problem ? I've been having observing the same problem that my Yaw always gradually drifts back to South (probably my axis is reverse) after a very short while pointing to the right direction.

Hi

I have used the Pololu MinIMU-9 v2 sensor in the recent past. I did not experience any of the problems that you mention. Maybe the settings of the sensors are wrong.I posted a library for interfacing with the V2 (http://mbed.org/users/Euler/) with proper settings. This library works well. Also, with the magnetometer, be very careful that there is nothing of metal in the neighborhood. Where I live, the magnetic vector comes out of the ground. When I put even a small metal object (like a nail) next to the sensor, the readings of the magnetic sensor change substantially. My advice now is that it cannot be used with metals in the neighborhood. By the way, I switched over to a gyro-free IMU with the ADXL362. They are very cheap. On the internet, you can find a lot of information on how to do that.

about the link you just posted http://mbed.org/users/Euler/

i found the library but i didnt have an example ino file for it i dont know what to do ??

can you help me ?