I'm still reading everything I can find; haven't actually built/bought anything yet.
Am struck by what appears to be a central hysteresis trade-off: Big props are efficient, but not responsive; small props are responsive but not efficient. Obvious response seems to me to be an "asymmetric hex" design, with three small fans for aiming (aka "point") and three big fans for pushing ("or "shoot").
No one seems to be building this. Why?
- Is the hysteresis trade-off less important than it seems?
- Is it 'just' that the standard controllers don't support this configuration?
- Is there some problem with this idea that I don't see?
- Or have I maybe just maybe come up with a good new idea?
Replies
To be honest, some drawing would make it easier to understand.
If this is, what you mean, there is a problem, that the big props will push more,if you use a standard firmware.
Of course,it depends on the direction you are flying, but I don't think, that this layout is revolutionary :D
I was thinking of something a bit more symmetric:
Im guessing in theory this would work. But I think it would depend on how the software assigns power to each motor for lift. You might have to modify the %'s going to each motor. Otherwise I would think the APM would try to equalize lift across all motors, and the smaller ones would be at max, while the larger ones would be lowered to keep thrust equal. You could try and put smaller props on 3 motors, larger props on the other 3 and see what happens. Im sure a developer would have the answer to your software question though. Physics wise Im pretty sure it would work fine.
"Asymmetric hex"
ok, to be honest, that is even symmetric in some way.
But jokes aside, I still think there is no fitting firmware, that would be able to handle that thing.
The bigger props will always push more with a quad or tri firmware.
But to be honest,if you make one of the big props like the backprop of a tricopter, it should work.
All the little props will be supporting the big ones.
But I don't guarantee for anything.
Joel
Yes, I know. I thought of names like "paired triple" or "complementary triple" but figured they weren't really any clearer, and:
I did figure the asymmetry in the motors was most likely to cause firmware complications. Otoh, given all the other software we'd have to write for a robot that needs a particularly fast and agile platform, tweaking some open source control software would be among the least of our tasks.
What do you mean? A straight tri would have bigger props than this mixed model?
Yes, perhaps. But this robot 'needs' to be radially symmetric and not have any preferred orientation, at least until it has to roughly align its charging rails with the troughs on its charging station.
I really don't understand this pp! Seems completely backwards, to me: the big props are keeping it up and moving it fast; the little ones are only for aiming. (Unless a big motor dies, of course.)
Ok, I got your idea now.
But for this kind of setup you will need custom firmware.
But I still think,that if you are using only three props, you will need to make them be able to move.
Otherwise, it wont work properly/you will always need to do weird turns to bring it around.
At least, thats what I think.
I don't think the he would need weird turns. At worst he could offset/slant one motor to help neutralize the torque like a tri copter. Once balanced the normal reduction/addition of power normally used should work.