I just finished building a tilt-rotor quadcopter. I started the project about a year ago, starting with a 3DRobotics quad frame to learn the ropes. At this point, the only thing that remains from the original quad is the APM and ESCs.
I planned to add wings on the arms to allow the quad to generate lift while flying forward, but I haven’t gotten around to trying that yet. The original design also had four servos so each arm could tilt independently, but that ended up being too heavy and too difficult to synchronize the servos, so I ended up changing it to have just one servo in the front and one in the rear. This lets the quad fly forward/backwards without having to pitch forward and allows for differential control of the two servos to control pitch in forward flight.
The main drawback right now is that it's too heavy. For the next version, I'd like to reduce weight and streamline the design. I'll probably replace the wooden frame with carbon-fiber, get rid of the wheels, and use the space inside more efficiently so I can fit the APM and an FPV system inside the body. Below are some videos and pictures from my build. This is my first blog post on diydrones, so I hope you like it!
Here's what's on the inside:
Completed iQuad before flight:
(the mast on the back was a GoPro mount which I ended up removing to reduce weight)
Comments
Hey, Im a senior in engineering, working on a project very similar to this. Need to get it working for in order to graduate, so its crunch mode. The method of controls that you described :
Could you explain a little more and maybe post that section of coding? My group is making a VTOL tilt rotor airplane, but instead of having control surfaces for yaw pitch and roll control, we plan to change the angle of the motor (via adjustments by servos on each mount) when the drone is flying in full horizontal mode. But during the transitional phase as we tilt the rotor/nacelles forward, the idea is to have the users input for roll, pitch, yaw be translated as a combination of throttle control (quadcopter mode) and thrust vectoring by changing the angle of the motors (plane mode).
Example:
User Input: Roll 15 deg to the left.
Quad Copter Mode (motors at 90deg up)
==> The two right motor increase throttle by 10% , the two left motors decrease throttle 10%
Transitional Phase (motors at 45 deg)
==> The two right motor increase throttle by 5% , the two left motors decrease throttle 5%
AND
==> The two right motors tilt up 2deg from set position (47deg now)
==> The two left motors tilt down 2deg from set position (43deg now)
Airplane Mode (motors at 0 deg)
==> The two right motors tilt up 4deg from set position (4 deg now)
==> The two left motors tilt down 4deg from set position (-4 deg now)
Does that make sense? Have any ideas of how to go about coding for this? Any help would be much appreciated!
Thanks!
I haven't posted the code yet... not familiar enough with github and with the community to know what the best way of doing that is. Maybe once I finish the final step of this project (which is adding wings), I'll look into how to post the code.
Hey there
Nice little thing you going on there. I was wondering what's the torque rating for the servos you are using, and how did you came to choose that rating?
Keep on the good work!
Cheers
Id love to build this, is the code available?
Also, for the new, lighter weight version that I'm building now, I'm getting some custom aluminum parts machined to reduce weight, but I'm still planning to use the pillow blocks as the bearing. I'll post photos of my new setup soon.
I got most of the parts for the rotation mechanism from Servo City (servocity.com). Just search for "Pillow Blocks" on their site.
Hey, looks good. Just wondering what you used for the bearings? Did you make the housings for them or did you get them from somewhere?
That was one of my initial concerns, but the stock APM software seems to handle things ok at low angles, and I can't get anywhere close to 45 degrees because of how heavy the whole thing is right now (at least I haven't been brave enough to try). Also, the controls for yaw and roll basically switch as the rotors pitch forward, but that should all be fixable in the APM software. The biggest control problem I had so far was the whole drone pitching back as I tilted the rotors forward which I fixed by getting rid of extra weight, setting the balance slightly forward and increasing the pitch stabilization constant. What I realized after some trial and error was that I needed different constants for roll and pitch because the weight is primarily distributed along the centerline instead of at the center.
That's pretty cool. Do you have any weird roll / pitch coupling? If you pitch forward at 45° and try to roll, how does it behave?