Using existing control arms from a 1/10th~1/5th scale surface vehicle to support the motor. Mounted ~90° from intended for surface. Yaw could be controlled with a driveshaft. The servo shown would shift the thrust away from CG, allowing balance corrections with out the need for more motors.
Of course a host of calculations for proper angles of motion, as well as coupling motions with hardware and software would be needed. I can understand most, but do not have the background to make solid assessments. Lots of trial and error to come! Gyroscopic forces involved are definitely something I do not fully grasp.
The intention is to create a two rotor that can be flown with a great deal of confidence. Then look at furthering the design into one that would be able to transition in to forward flight. This would mean a drastic look at software as yaw and roll control would swap depending on mode of flight, and need some sort of mixing in transition. All stuff for later down the road...
I plan on starting with a basic tri just to get the yaw mechanics sorted out. Gonna go to the hobby shop tomorrow and pick out the plastic parts. CNC'd 7075 and G-10 will likely make up most of the remaining parts. Hoping that in borrowing from popular surface vehicles this project can be cheap enough that some might follow along.
This is an awesome idea. Something similar has been implemented before (or at least prototyped): http://www.youtube.com/watch?v=rK61Q0crRGo
It would be cool to see it done with APM though. Imagine an autonomous scorpion :)
I've found a few tries
that one does surprisingly well.
Wondering if shifting the battery or the motors would be a better idea.
Ended up here after buying some suspension parts...
Likely to shift towards a Tri or go straight to a twin before building this though.
I am working on a project that is based on APM1 (later probabilly APM2) and the configuration is similar to a twin copter (two motors and two servos). Unfortunately I have not the necessarry C++ programming skills to implement the modifications of the ArduCopter software. I have o cooperation with a SD but the guy has nothing to do with our hobby. The idea is to make a library to support this configuration. My questions are: Do you think is this possible (I mean solve the problem with one library) and if I send the modifications can one of the developpers look at them and maybe help me out. This will happen in the next few weeks.