RC Groups

RC Groups
    Indoor Pattern/F3P
        Discussion Another coaxial F3P plane

#1 Kimmo Kaukoranta Feb 15, 2011 07:43 AM

Another coaxial F3P plane
2 Attachment(s)
I was going to keep this under wraps until the Lithuanian championships, but since Fabien already told about his system in the spies thread, I thought I might as well post something on a bit different coaxial motor system I have been flying since last december.

This one is using 2 AP-03 outrunner motors weighting 3.25g a piece. Since there is no way to run a 10" prop direct drive with a 3g motor, gears are used. To give You an idea of the size, the diameter of the motor is around the same as the hub in a 10" GWS prop... pretty small.

The motor/gearbox system weights around 16g and the 10" carbon fiber props 2x3.5=7g. I'm using two YGE-4S speed controllers (2x0.7g=1.4g without wires) connected to different channels on the Spektrum AR6100 receiver. This way I can limit the throttle for the motor turning the front prop to match the two torques.

Flying weight of the Manta v3 with the system is 107g with a Hyperion 2S 120mAh battery. Full throttle current is around 3.6A (the whole system). As for the efficiency, with the 120mAh battery You can get trough 2 complete sequences, but not much more. Pretty OK in my opinion.

Suomen CUP 3/4 Laukaa. F3P-F11 Kimmo Kaukoranta (2 min 33 sec)

Suomen CUP 3/4 Laukaa. F3P-P11 Kimmo Kaukoranta (2 min 34 sec)

#2 wparsons Feb 16, 2011 07:31 AM

Very cool! It seems to climb slowly very effortlessly without the prop torque!

#3 3Don Feb 17, 2011 09:02 AM

That is nice! Looking forward to see it in Lithuania.


#4 Kimmo Kaukoranta Feb 20, 2011 10:49 AM

6 Attachment(s)
Just finished another Manta with a coaxial power system. This one is cut open and covered with 5 micron mylar on both sides instead of milling. Pair of 10" props 5.8g total, 20.6g for the whole motor system including the props and 98.5g flying weight with a 2S 120mAh battery. Flies even better than the first one :)

Also made some modifications to the older one. Shortening the wires etc brought the weight down to 105.8 from 107g. Getting rid of the wing and elevator tip fenches saved another 1.9g, should be right around 104g now.

#5 skirmish Feb 22, 2011 07:22 AM

Excellent work. True aeromodelling!

#6 Boogie3D Mar 27, 2011 03:29 PM

What is the gear ratio of those?
Did you just make your own "left" GWS prop to make a mould for those props used?
And why is the distance between the props so large? I will ask my friend who designs coaxial props if the smaller distance wouldn't be better. I am only working on single prop design right now.

#7 Mike Seale Mar 27, 2011 04:11 PM


Originally Posted by Kimmo Kaukoranta (Post 17462821)
This one is cut open and covered with 5 micron mylar on both sides instead of milling.

A couple of years ago someone posted a report that a milled model with weight added to make it the same as an unmilled one still flew better. Reasoning was that the milling adds drag to slow the model down. Would be interesting for you to try the same with mylar one side and then both and see if there is any difference in the way they fly.

Very neat looking model, btw.


#8 Kimmo Kaukoranta Mar 28, 2011 06:59 AM

skirmish, thank you very much!

Boogie3D, the front prop is 7:1 and the rear is 5.83:1. This allows the rear prop to run a bit faster to compensate for it being in the wake of the front prop. The prop molds are made from commercially available props that are available as a contra rotating pair.

About the distance of the props. At first it was done based on what looked about right. Later I read a paper on coaxial micro air vehicles where they had actually studied this. They concluded that a distance of 0.35x the radius of the props is about right. Anything smaller than that will affect the efficiency. I was lucky to hit just around that figure :cool:

But they also concluded that although the efficiency will suffer, the difference is not huge. So it could be possible to shorten the distance a bit, but not too much. The amount of interference noise will probably increase. Also keep in mind You do not wan't the blades hitting each other in a tail slide or similar maneuvre :rolleyes:

I will add a link to the study if I can find it. Edit: Here

Mike, I did some test pieces and the decision to cover both sides was based on the added stiffness. I mounted the mylar on a wooden frame, used a heat gun to tighten it and glued with 3M 77. This yelds a much stiffer part than milling or covering just one side. Covering one side would obviously be lighter. The two 10" props with the gearbox offer plenty of drag on idle.

#9 Mike Seale Mar 28, 2011 08:20 AM

Thanks, Kimmo. And adding mylar to both sides also reduced / eliminates warping.


#10 David Lundstrom Mar 28, 2011 03:28 PM

Hi Kimmo. Impressive work!

How much thrust are you getting out of your contra system at full power (3,6A) ?

I’ve also been playing around with a custom contra system this winter. My system is based on two outrunners similar to turpaud. I’ve been pretty serious with testing different windings, props, and dyno measurement of the individual motors. Result is a 28g motor that turns two APC 10x3,8 @ 7A. Total thrust of 270g. My conclusion with contra props is that flying characteristics are lovely, but 28g is too heavy for F3P.

I made my motor out of curiosity to fly something with contra props. Since I like it so much I’m considering making a smaller version more optimized in weight. I figure I can do a 20g motor with performance suitable for F3P. A geared solution like yours however will be more efficient and even less weight (as you have shown). The question is how much the difference would be between two equal systems.

See you in Jämi by the way!


#11 DaveL322 Mar 28, 2011 07:55 PM

1 Attachment(s)
David you can see Kimmo system on my ETOC Symmetric, my plane is heavier than plane Kimmo use, but for European sequence it is still ok. However for ETOC sequence i would like to have more power. I compensate lack of power with biugger battery 250mah 2s, and beffore flight i heat it, result is good enought for almost constant speed.

DonatasPauzuolisSymmetricETOC (4 min 9 sec)

Donatas Pauzuolis

#12 Kimmo Kaukoranta Mar 29, 2011 01:59 AM

Hi David,

I actually have not measured the static thrust of the whole system. I did however test the single motor on a test stand and the figures are 7.4V (lab power supply), 1.9A, 14W, 133g using the 7:1 gearbox and 10x4.7 GWS SF. With the 5.86:1 ratio and same voltage it's 2.3A, 17W and 139g.

For the rear prop those figures are not valid since it's in the wake of the front prop. It will use less current. I limit the throttle channel for the front prop so the torques are equal in a hover or a slow vertical upline. I believe Donatas is now not using that limitation because of the higher power required for the symmetric (it is a bit larger model than the Manta) and the demanding EToc schedule.

To make the system work even better in a model >120g, I think I will need to increase the amount of built in compensation (now there is only the small difference in the gear ratios). That would mean increasing the pitch of the rear prop, a lower gear ratio for the rear prop or a higher kv motor for the rear. That would remove the need for limiting the throttle on the front prop and increase the total power available without increasing the weight of the unit. Also Donatas would not have to heat his batteries anymore :) For a lighter model I think the system is fine just as it is.

I think a geared motor is always more efficient when trying to run a large prop with a very small and light motor. It will also brake much better with the same prop as an direct drive outrunner. I think it also has a faster acceleration.

Actually I have first built several single prop geared systems before the geared coaxial due to those benefits. The latest one, which also Janne Lappi was using in Lithuania weights <11g and provides 210g static with 7.4V 3.4A 25W input power using a 10" prop. I do not think there is an direct drive motor which can match that. If anyone is interested on that I can write a bit more on that in a separate thread when I have the time.

By the way. It is possible to build a coaxial using a single brushless ESC, single outrunner motor and no gears at all. You do not need 2. Together with Seppo Pekkala we have built 3 of those :rolleyes: Who will be the first one in this thread to figure out how that works?

#13 Kimmo Kaukoranta Mar 29, 2011 07:51 AM

4 Attachment(s)
Here are a few photos to show the size of the motors in the coaxial unit. Axi 2203 and my fingers as a size reference :)

#14 David Lundstrom Mar 29, 2011 10:24 AM


Originally Posted by Kimmo Kaukoranta (Post 17818062)
Hi David,
By the way. It is possible to build a coaxial using a single brushless ESC, single outrunner motor and no gears at all. You do not need 2. Together with Seppo Pekkala we have built 3 of those :rolleyes: Who will be the first one in this thread to figure out how that works?

Well, you did not say the outrunner were really brushless, so I figure its a brushed brushless outruner :rolleyes:, meaning some sort of brushes or slipring is used to feed a "counter spinning stator".

If that is the case why did you abandon that solution? Sounds like a really clever idea.


#15 Kimmo Kaukoranta Mar 29, 2011 10:51 AM

Yes, that is correct!

You take something like an Axi (we used the cheap copy from Turnigy for this), replace the shaft with a longer one and mount the other prop to the end of the shaft. The stator, the part that normally is bolted to the plane, is hang from bearings to allow it to rotate to the other direction and the other prop is mounted on that. Now You have those 3 wires rotating with the stator, so 3 slip rings are used to connect them to the ESC.

A few interesting questions arise from this:
-If both parts of the motor rotate, is it an inrunner, outrunner or both?
-It uses a brushless ESC but has brushes. However the brushes are not used for commutation. Is it a brushed or a brushless motor?
-The part where the windings are is normally called a stator. Now it is rotating as well. Is it now a rotor? Which part of the motor is the stator?

As crazy as it sounds, it actually works and very well. I think the system has some potential to be developed further, I have not completely abandoned it.

I went with the geared solution on this one mainly because I figured it could be done lighter. To make the slip rings smaller and lighter would require some special tooling. I also figured the geared solution would have better braking and faster throttle response, like the single prop geared systems I had built. But it certainly has some benefits that could be explored further.

All times are GMT -5. The time now is 07:13 PM.