Ardu-copter flipping sideways instead of taking off
0Hello!
i have built an Ardu-quad-copter around the HobbyKing PixFalcon flight controller.
i'v made a serious effort to try to learn how to calibrate the drone using Mission Planner and tried all possible CW and CCW combinations - Yet the quad-copter just stays in place or flips on its side - instead of taking off symmetrically.
Can anyone please help me solve this problem?
Thanks.
- 9 comments, 7 replies
- Comment
Make sure the props are on the right motors and the motors are spinning in the right direction.
@roineust, +1 what DJMajickMan said. Sounds like you've got motors spinning in the wrong directions, or rotors on the wrong motors (ie: CCW rotor on a CW motor). The placement of CW and CCW motors also makes a big difference to MP. You must have motors spinning like this illustration:
And of course you'll need to have the correct rotors on the corresponding motors.
By the way can you upload a photo of it? Include the props on it and the direction they are spinning if possible :)
One of the motors strats spinning later than the others, could that indicate about the reason? i found this video, that explains how to calibrate ESC's one by one, could that be the solution?
@roineust ESC calibration is an imperative part of the FC setup. Did you use the Mission Planner setup wizard? ESC calibration is covered in the wizard. But you can also do a manual calibration process depending on which firmware your ESC is using.
OK -
Now that i'v gone through ESC firmware upload again and calibrated the ESC's - all done on the ESC's one by one - i'm kind of lost, since i still get a sideways flipping, instead of a symmetrical take off.
b.t.w the motors all start and speed up in sync now - so that is probably not the problem...
Any ideas what should be next on my problem elimination check list?
Thanks.
@roineust Can U post a short (maybe 30 second) video of this phenomenon? Ideally with the drone in full camera frame...
Sure - here is a video i just took. You can see i am pushing the throttle stick up 3 times - and letting it back down every time - since i already know it ends up in a flip at the direction it starts tilting in the video. That is, i don't want to push the throttle any further and damage the props further than already damaged in previous 5-6 full flips:
http://sendvid.com/4qy53uuc
@roineust Something else to check is the signal from the FC. Make sure it is on a flat level surface and reports back to you that it is. Can you also get a closeup shot of the props and motors at a good angle to check orientation.
Here is the FC i have:
PixFalcon Micro PX4 Autopilot
I don't seem to find decent documentation.
When i connect the 4 motors to channels 1234, where channels 1 is a Futaba with black, red, white wires and channel 4 is blue wire Futaba with only data and channel 5 is yellow wire with only data and channel 6 is green wire with only data - i get 3 motors turning CW and only 1 turning CCW.
At last got it right! it's flying! But not yet very precisely..
So first of all: DJMajickMan, ruouttaurmind and anyone else who replied- Thanks a lot!
Now for the fine tuning issues: need it to take off as symmetrically as possible and to hover over a position, as stable as possible - any advice?
Need a 1-1.5 kg payload power and no need for speed or acceleration at all - Only stability at a 5 meter hovering height is what's important here.
Any specific type of frame, motors or props to recommend, in relation to such a mission?
Thanks!
@roineust Glad you got it sorted! What did you find that was causing the problem?
RE: Documentation for your PixFalcon, I'm guessing it's a clone of the 3DR PixHawk, yes? If so, the 3DR documentation might help put you on the right track.
Stability with a 3.5lb payload, plus your gross hardware weight... I'd probably step up to a hex or octo configuration.
Couple questions:
Do you have a requirement for maximum physical size?
Are you planning on any specific battery configuration?
Which GPS are you planning on adding?
With that payload, you'll want larger rotors driven by slower motors. 620's pushing 12 inch rotors, or even 320's pushing 14 to 16 inch rotors. More motors = greater payload and more stability, but also = more expense and bigger battery requirement.
EDIT: Wider is better as far as the chassis is concerned. You'll have better stability with a 1000mm wingspan than with a 550 wingspan for example. Notice how many of the pro cinematography drones are 1000mm or more.
Are there any other specific requirements? For example, retractable landing gear, folding chassis for ease of transport, etc?
This drone will have to hover over a fixed low position, as long as possible, to have a good payload ability (1kg+) and to be as compact a possible. In what way, would you say a 3 blade prop could be compared to a 2 blade prop? Could such a comparison be done? Fore example, is it correct in very general terms, to say something like: A 10'' 3 blade prop has the same ability as a 12'' 2 blade prop? If it is generally true to say such a thing, does a 4 blade prop exist as well as an over the shelf product and does it make it even better for such a mission? How low should one go with the motor KV, if all needed is max hovering time on a fixed low position and that mentioned payload ability? Is there any other type of component in such a system, that has certain properties chosen, that could be essential for such a mission?
@roineust I think what you're looking for is a pole. Infinite hold time with very low power consumption and high payload capacity.
Jk. Every year HobbyKing has a heavy lift competition that you could probably check out some past entries for ideas.
3 bladed props of the same diameter draw more power and produce an increased amount of lift, but i think the power draw is greater than the lift increase. Check out a power calculator like http://www.ecalc.ch/motorcalc.php to run numbers on kv, voltage, prop diameter and pitch and give you thrust and runtime.
This is a multicopter specific calculator: http://www.ecalc.ch/xcoptercalc.php
Pole..ha! thanks djslack, i'll check out the ecalc.