Replies to This Discussion

Permalink Reply by Ricardo Madera on July 27, 2010 at 11:50am

Interesting, What ESC's are these? Im working on something similar to you XBees except that im using SunSpots instead. Their purpose would be to ease the configuration process by doing it wirelessly, and to run some vision algorithms using a cmos cam for a obstacle/edge detection. I am also using ArduCopter but i wrote my own code from scratch (except for the accel/gyro filters). Im working on a configurator for it now. Im also planning to do a sliding mode controller with back stepping to have really good stableization. I just need to decide on which is the best way to measure the propeller RPM's.

• ▶ Reply

Permalink Reply by Andrew Robinson on July 27, 2010 at 12:09pm

You might want to look into the ESCs. The ones I use are TowerPro 25As, modified to be I2C capable, but I wouldn't wish that modification process on anyone. A fellow around here sells modified ESCs for some mark up, you might want to consider purchasing his.

Sliding mode control sounds like an interesting approach. Please keep me updated on your efforts and theory.

I believe that you should be able to measure RPM by taking a ADC reading off of the unused coil. You would need to dive into assembler to modify the ESC's programming however, I don't think any of the current implementations would suffice. Another method would be basically a tachometer. You could easily reflect light off of your props.

XBee 900Mhz are cheap and can provide buad rates up to 115200kbps. Good enough for me! I'd like to put a high speed serial link on the copter (wired 1mbps over emulated UART possibly) and see about offloading the algorithms to MatLab for development. I'm not sure if the latency induced by the XBee transmission protocol will be acceptable for real-time control.

• ▶ Reply

• ‹ Previous
• 1
• 2
• Next ›
• Page