So for the past two months I've been working on a quadrotor project, and I've designed an ARM based board just for it. Contains a TI LM3S9B96 ARM Cortex-M3 (not out yet, some how convinced TI to send me a few), 3-axis gyro, 3-axis accel, compass, zigbee, 4 motor pwm outputs and temperature sensors (also tons of breakout pins). I plan to make a shield with gps, barometer and a couple of other things.. Just now I have confirmed the board completely works, and I'm beginning to get into the kalman filter and pid control loop.
I'm just wondering how many people would be interested in the board. I'm likely to make it open source hardware/software.
Link to the website: http://wiki.032.la/nsl/Quadrotor
Juan! THIS LOOKS AWESOME!! Nice work man!
Where did you get the chassis? I've ended up designing my own chassis out of very lightweight aluminum, and it's out currently getting machined. I ordered 5 of my "drone" board, and I hope to see the PCB's shipped in the next few days. So there are some exciting things to coming soon.
With all this new hardware, we both have some PID tuning to do!
What Kalman Filter you are using?
(9-DOF) MPU-60XX + Cortex M4 would be pretty interesting combination...
Got my new board!
Soldered it all up and it worked the first time :)
So now it's time to refine the code and assemble the new chassis coming in next week. I can finally say it's a quadrotor when it's finished milling. Once the chassis and board+other components are together, I'm going to build a test apparatus to do a test flight.
I'm programming the microcontroller via JTAG right now, but I'm going to write a bootloader once I'm less occupied with other things. I'm using the IAR environment and uploading/debugging with it.
It's a custom made Kalman Filter, nothing too fancy. I'm going to release the code in July along with a power point explaining all the code (probably make a video on it too).
That's a neat sensor you mentioned... I think I might use it in a later board :) Opens up a lot of real estate.
I have tried to find example code of sigma-point kalman filter with poor results...
Offtopic: Iv found that the maximum sample rate is 200 Hz if you are using quaternions or rotation matrix with sensor fusion. You can still read data from gyro/acc with max sample rate. I think that 200 Hz would be enough for orientation estimation, but it is not prolly enough for PID controller. I was wondering if it is possible to calibrate gyro data with sensor fusion output and use that for proportional component (gyro drift cancelling). I guess that gyro data could be used for derivative component without calibration (drift does not have effect).
The picture is a bit out of date, but the quadrotors IMU and control code is complete. All that's left now is improving hardware to reduced high frequency vibrations from the motors, as well as new props. The control board is going through it's final redesign with tons of new features and hardware. My target price tag is $150 so hopefully within the next month or two we should see the board hit the market (I will provide details very soon!!). Within the next week I plan to be properly balancing on 3 axis (roll,pitch,yaw).