Posted by Dan Overholt on August 10, 2009 at 3:16am
Hi Bill,I've followed your UAVdevboard developments for awhile now, and have been very impressed! I don't yet have one of your boards (or a plane for that matter!), but am thinking about getting into it soon.What I have in mind, at least after getting up to speed with your current developments with AileronAssist (possibly with a Multiplex Gemini biplane a friend of mine might sell to me - do you think this would be an OK choice?), is to build a new type of flight controller that would plug into the trainer port on the back of the transmitter. Imagine a kid holding a small toy plane and "pretend flying" with it - what I'm thinking of doing is allowing the motions of pretend flying to control the real RC-plane through a sensor-equipped "toy plane" and your UAVdevboard.Obviously, this would require 2 of your UAVdevboards, the one in the RC-plane, and one in the "toy-plane-controller" (or I might just use my own board in the toyplane controller to start with http://create.ucsb.edu/~dano/CUI/ and a 3-axis accelerometer, which should be good enough to start with in terms of roll/pitch orientation determination) -- correct me if I'm wrong, but as I understand your DCM algorithm, since you say the plane will return to level if you let go of the sticks, I am assuming that if I hold the control stick at a 45degree bank angle, that the plane will also hold the 45degree bank? This would make it quite simple to do a quick prototype of my idea, where the rudder/throttle are still left under control of the transmitter's left stick, and the ailerons/elevator are controlled by the "toy plane" (basically just replacing the right stick). Hmm, I just realized a basic problem is that the trainer port would not allow "partial control" like this - either total external control or total internal (not 1 stick internal and 1 stick external) ... well, I could just open up the radio and hack into the circuit with a DPDT switch for control of the right stick...I might also make the toy plane wireless eventually to free it up from cable tangles being plugged into the transmitter's trainer port (a couple of XBees). And the logical end goal - although I know this would require magnetometers on both the UAVdevboard and in the "toy plane" more like the PICpilot (actually, the newest iPhone has a 3-axis magnetometer and 3-axis acceleromoeter, I might use that, but of course instantaneous response is sacrificed a bit without any gyros) - would be to incorporate rudder control as well and allow full aerobatics all controlled with the toyplane controller (e.g. if you hold the toyplane vertically, the RC-plane should hover, if you hold the toyplane upside down the RC-plane should fly inverted).I realize this got a bit long winded, but I'd love to hear your thoughts on this, and just wanted to compliment you on your hard work and sharing of the efforts!Best regards,Dan
You need to be a member of diydrones to add comments!
I have been thinking about this as well dan. I currently have a wii remote in bits in my workshop and have been looking at hopw to control the aircraft with it. I have managed to get it to control my pan/tilt camera mount but that is just basic and using the same principle as the headtracker.
You may or may not need a magnetometer on the flight controller. It depends on how you define the interface. I think that there is an intuitive way to define the interface such that the yaw orientation of the hand-held is irrelevant, so you would not need a magnetometer or IR on the hand held. All you would need is gyros and accelerometers. Aside from interfacing issues, the UAV DevBoard would do the job.
Also, there is something called "DCM lite" that will give you two axis measurement, without needing to know anything about yaw. So, what you could do would be to roll the hand held to make a turn, and pitch it to go up and down. That is all that is really needed. On the aircraft end of the control, these two signals are all you need to command rudder, elevator, and ailerons.
Interestingly enough, when I was about 8 years old, I had a toy electric car that had a wired remote control that worked the very way that I am suggesting. You held the control level and rolled the handheld left or right to make a turn. The way they designed it (I eventually took it apart, that is what geeks do) was there were metal balls that rolled onto resistive strips. It was very intuitive and easy.
I think the "controller" side hardware could be simpler than the airborne side, since it's stationary. Basically you just need a wiimote-style sensor setup, right? How about using an actual wiimote?
Very interesting idea. Technically, it could be done, I think, it would even be possible to implement a reasonable interface without magnetometers.
I have been flying for 25 years using joysticks, and I am not very good playing Wii games, so I am probably not the right person to give you any useful feedback on the new control interface that you are suggesting.
It will be interesting to see what other pilots have to say.
Replies
You may or may not need a magnetometer on the flight controller. It depends on how you define the interface. I think that there is an intuitive way to define the interface such that the yaw orientation of the hand-held is irrelevant, so you would not need a magnetometer or IR on the hand held. All you would need is gyros and accelerometers. Aside from interfacing issues, the UAV DevBoard would do the job.
Also, there is something called "DCM lite" that will give you two axis measurement, without needing to know anything about yaw. So, what you could do would be to roll the hand held to make a turn, and pitch it to go up and down. That is all that is really needed. On the aircraft end of the control, these two signals are all you need to command rudder, elevator, and ailerons.
Interestingly enough, when I was about 8 years old, I had a toy electric car that had a wired remote control that worked the very way that I am suggesting. You held the control level and rolled the handheld left or right to make a turn. The way they designed it (I eventually took it apart, that is what geeks do) was there were metal balls that rolled onto resistive strips. It was very intuitive and easy.
Best regards,
Bill
I think the "controller" side hardware could be simpler than the airborne side, since it's stationary. Basically you just need a wiimote-style sensor setup, right? How about using an actual wiimote?
Very interesting idea. Technically, it could be done, I think, it would even be possible to implement a reasonable interface without magnetometers.
I have been flying for 25 years using joysticks, and I am not very good playing Wii games, so I am probably not the right person to give you any useful feedback on the new control interface that you are suggesting.
It will be interesting to see what other pilots have to say.
Best regards,
Bill