Y3 Tri-Copter - with ball bearing yaw mechanism (WIP)

[LennyPane]

Hey Guys,
i'm new here, and starting to plan my first build with a tri-copter design. love the flite test videos, very good to get to grips with everything!

my plan is, i want it to be cheap (around 170EUR/200$), but still be accurate and have the capacity to race or start to learn aerobatics.
i already did a lot of research, got confused, then less confused, then more confused(mainly about motor, prop and battery sizes)
i'm just wondering if my plans and parts will cooperate and fit together. but im looking forward to fiddling and figuring it all out as i go.

//scroll down for pictures//

PART I
15.1.2015

My ideas/plans:

The Tail Yaw Mechanism TYM
the 10x10mm tail arm is inserted into two 10mmØ ball bearings, rotating the whole back arm, instead of just the motor. and fastening the arm and ball bearings into the stiffened tail end of the main frame, hopefully making it really robust, but also freely tilt-able with no resistance from.

and not only that, but with this config. i can put the yaw servo right in the middle of the copter, behind the arm inside the main frame thus keeping all the parts in the middle and the center of gravity focussed and centered.

Center of Gravity
i will also try and mount the motors on the ends of the arms as well as 'bend' the arms down if i manage to get that done. all that to keep the center of gravity (height wise) right in the middle.

(instead of the main frame 'dangling' under the motors like a shallow pendulum... but.. i guess moving the arms and motors down like this will actually bring the CG below the main frame because the force is transmitted at the bottom of the motor where it is fastened to the arm, right?.. hmm not so good. although with the battery pack its probably spot on )

i also thought, maybe i can even mount the ESC's around the tail-ball bearing mechanism(white boxes, red cables - as pictured) and feed the motor wires through the hollow arms. keepin' it tidy
there's also a slot in the top plate under the KK1.2.5 board to feed cables through.

Planned Parts/Specs:
KK2.1.5 + 19S firmware
EMAX CF2822 1200Kv
ESC Simonk 30A
EMAX ES08MD 12g/ 2.4kg/ .08 sec
3000MAh Fleureon 30C, 11,1v
9x5 or 10x4.5 Props
Flysky FS-T6-RB6 FS 2.4GHz
maybe fpv; SKY2 800TVL (flat spot on the main frame front for camera)

Materials:
carbon arms, carbon frame plates, plastic frame fillers

Sketches: (made with blender)




tail yaw mechanism and ESC to Motor cables


folded arms


inside main frame without servo - ball bearing yaw mechanism, ESC to motor cables
_________________________________________________________





Planned Dimensions:


everything is not fixed yet, don't know if 3D dimensions correspond to the real world...
all the bought parts(esc's, KK2.1.5, motors etc.) have the correct proportions in these sketches though

if you have suggestions, critique, ideas, words, unicorns feel free to share.
cheers,
Lenny

 

[saiga556]

I like how you decide to make the whole tail boom rotate. Seems like it would keep weight near the center of the craft. Also seems like the yaw mechanism is more protected where you have decided to put it.

Man I wanna build a tricopter bad but the cost of fpv gear has stood in my way.

 

[LennyPane]

I like how you decide to make the whole tail boom rotate. Seems like it would keep weight near the center of the craft. Also seems like the yaw mechanism is more protected where you have decided to put it.

Man I wanna build a tricopter bad but the cost of fpv gear has stood in my way.

exactly, thanks! trying to make it as maneuverable as possible.
yeah my empty pockets are keeping me from building it right now. and fpv gear is so friggin expensive..

 

[johnmw]

interesting concept...
while having bearings minimize friction,
rotating the tail boom means heavier load to servo (compare to conventional).
not sure if it will increase delay in yaw response or otherwise negligible?

 

[LennyPane]

interesting concept...
while having bearings minimize friction,
rotating the tail boom means heavier load to servo (compare to conventional).
not sure if it will increase delay in yaw response or otherwise negligible?

the tail boom rests inside two relatively strong bearings with the main frame supporting it, so my guess is there will be no load at all on the servo because the tail just sits there fixed in place except for the rotation, and then the servo does its job on top of that...
or do you mean because the boom is 'bent'? thought of making it straight

 

[makattack]

johnmw is referring to the loads as the servo spins the whole assembly with motor and stops it. Now imagine a flight controller that's controlling it for you to adjust for yaw dynamically as wind/etc affects it. There's normally a very high load on a tricopter yaw/tilt servo, so this might be a lot. Also, the weight of large bearings is something I'd be concerned with. Lighter = more flight time and less damage on crashes!

Still, it's a very interesting design. I suspect the other difficulty is in getting bent carbon fiber arms... which would also affect the torque on the tail server as well, unless you can position a motor in such a way that it's perfectly balanced along with the prop and accessories on the tail boom.

 

[LennyPane]

johnmw is referring to the loads as the servo spins the whole assembly with motor and stops it. Now imagine a flight controller that's controlling it for you to adjust for yaw dynamically as wind/etc affects it. There's normally a very high load on a tricopter yaw/tilt servo, so this might be a lot. Also, the weight of large bearings is something I'd be concerned with. Lighter = more flight time and less damage on crashes!

Still, it's a very interesting design. I suspect the other difficulty is in getting bent carbon fiber arms... which would also affect the torque on the tail server as well, unless you can position a motor in such a way that it's perfectly balanced along with the prop and accessories on the tail boom.

right, very good points sir, thanks!
i will have to see about the wind/etc effects on the tail and the load, but i thought since my servo supports 2.9kg that would be enough.. maybe i need to go for an even bigger one theres still space inside the frame.
and i found some 10x19x7mm bearings, hoping they are light since they are small.

that was my idea as well, to balance the arms exactly so they have similiar characteristics to straight ones for this purpose(really tapping in the dark here, have to try this out with wood or something)
about bending them... i thought cutting them up in precise angles and bolting them together again. but you know, as im thinking about it... it's probably not worth it to sacrifice boom integrity, servo torque and what not... for that 1-2cm elevation.. de-elevation

 

[saiga556]

Maybe you could build your booms out of cf sheet. Make them with 2 pieces that are your side profile and a piece running down the center that forms a sort of bent I-beam. That would also give you a channel to run your wires in.

I say build it with the bent arms and definitely go with the bearing idea. The best part about building yourself is ending up with something you can't get anywhere else.

 

[makattack]

Here's a fellow (from the UK) who made his own frame with curved arms and a torque tube mechanism using teflon bearings with a 10KG servo:

http://www.rcgroups.com/forums/showthread.php?t=1760084

It looks like a lot of work, and back when he did it, he obviously didn't have as good a selection of props, motors, escs, flight boards!

 

[LennyPane]

@saiga556 yeah exactly and great idea!! found some hollow square carbon tubes (1x1cm, 2mm thickness i think) on ebay, but with your idea i have a lot more control over the shape of the booms! plus, i can control exactly how it sits inside the two bearings.
just wonder if thats gonna be as rigid as machined ones.. you think i could bend the I-beam part at that angle? or maybe just cut it
and yeah im definitely going for the bearings, thats for sure! maybe im gonna bend the arms too, just to be the first one(*hopefully) who made this design work

@makattack awesome thanks! does look like a lot of work, but good to see it has been achieved before..
didnt know about the teflon bearings before, super slim! great advice

 

[roversgonemad]

The gyro effect of the spinning propeller would far exceed polar moment contribution from rotating the boom. I don't know that it would make too much difference. Plus, the offset in the boom puts the center of the plane of the spinning propeller closer to the pivot. It would actually be easier to rotate if the rotational pivot of the arm was on the same axis as the center of mass of the propeller. It really should rotate fairly easily, at the least, negligibly more resistance than rotating the motor and mount only.

 

[LennyPane]

Thread Moved!

thanks for the replies and input guys, helping out a lot!
this thread was moved to here though:
http://forum.flitetest.com/showthread.php?15164-y3-Tri-Copter-with-Tail-Yaw-Mechanism-Build-Log
sorry for the double post :/ quite new to this.

@roversgonemad prop gyro effect. good point! but thats what i was thinking about, is it better to put the center of the prop in line with the arm rotational pivot center or the mounting point where the motor is fastened the boom? because the weight gets distributed/transferred there