Simple PID implementation and test in hardware.

[vmayoral]

The idea is to have the quadrotor having a stable flight using simple PIDs.

Once this is achieved the idea is to manufacture the boards for you guys.

[vmayoral]

Hey guys!

Merry Christmas to everyone. I could find some time to solder and test some new ESC/brushless motors combinations and this is the result. The robot needs more thrust from the motors but it looks promising.

[bugobliterator]

greetings victor, Great video, but it surely needs work, i just wanted to know if the erle is modeled or simulated over matlab or any sim software. If not I've got a guy in my team here who is doing the same for our future quad, so he can have some things tested, this will surely give us a proper direction.

[vmayoral]

Hey Siddharth,

Thanks for the feedback. We didn't simulated it ourselves but we took the work of Menno Wierema as a starting point. It'll be great if you could support us on this matter. What would you need in order to simulate it?

[bugobliterator]

Hi victor, There are some coefficient which need to be calculated for simulation which can be derived from a good CAD model. so it would be nice if you can send us a cad file of erle's model.

[vmayoral]

Great! I'll get in touch with our industrial designer and i will provide it as soon as possible.

Keep in touch.

[vmayoral]

@bugobliterator i'm sending you an email with the STEP and STL models of the board. I would suggest to create a BeaglePilot/simulation repo where we could push this sort of stuff. It'll be quite nice as well if you guys could push the MATLAB files so that we can test it as well.

[bugobliterator]

Sure, I'll do that and happy new year btw:):):)

[vmayoral]

Yes, indeed, Happy New year guys! Glad to be working with you on this project ;).

Best wishes for the year 2014.

[vmayoral]

@bugobliterator any news regarding the CAD model? Is what i sent you good enough to make the calculations?

[bugobliterator]

Sorry!! that issue skipped my mind. Yeah, i received the model, it was unable to provide us enough factors, there are almost 21 parameters that are needed for the simulation, so better idea will be that first we complete the simulation of the model in hand that will take no more than one or two weeks and than send you the step by step details/requirements for calculating all these factors so that you can send us those parameters itself, because there are many factors which can't be defined by the model e.g. the most basic ones such as mass and critical ones such as propeller and motor parameters.

[vmayoral]

Thanks @bugobliterator :). I created #6 for this purpose. Please keep us posted about it.

[vmayoral]

simple userspace autopilot implemented here. Performance bad, moved to RT_PREEMPT with the ArduPilot code.