I have taken a Smartdrones 4404 motor apart today and rewound it.

This motor is completely identical to the RCtimer 5010, 360KV motor, but it is more correctly labeled 4404, since that is the stator size.

http://www.rctimer.com/index.php?gOo...5&productname=

Here is the data:

The original motor was rated 370KV and it was wound with 4 strands of 0.25 mm, 40 turns per tooth. 370cm wire per phase.

Its a 12n14p motor.

I rewound it with 2 strands of 0.16mm, 100 turns per tooth. That equates to about 950cm per phase giving a resistance of 7.8 ohms across 2 phases.

I could do 20 turns per winding layer, but its tideous..

The resulting motor is very smooth, and has a lot of torque, but the gimbal its intended for (EOS M/Nex5 and similar) is not completed yet, but I am pretty sure it will do a good job of handling it.

Originally Posted by jazzboy

Oops, here it is:

http://www.rctimer.com/index.php?gOo...5&productname=

Andy

I'd say there almost certainly has to be. On the 28mm stator I wound I got 28 turns on layer one to make a full arm. Those were cheek to jowl and touching each other. When I later removed that and measured length, it turned out that the 28 turns worked out to 0.61"/15.5mm per turn with 32 AWG/0.18mm wire.

So on an arm like that, with about 5.04mm of windable length (28 turns of 0.18mm wire side by side = 5.04mm) the 0.20mm wire would work out to about 25 turns per layer and 80 turns would have fit easily.

If we all keep some notes and post the info here on the turn counts and length per turn it will really make this all go a lot easier.

When you wind, in a perfect world, all turns in all layers will lay side by side and there will be no crossing turns. But I've tried to wind two layers on the 28mm stator arm that way and it is very difficult to keep the turns that way. Part of it is my old hands, brain, eyes, and keeping track of turns. Maybe you youngsters will find it easier.

It is best to finish the last turn near the bottom of the "V" to keep the transit runs closer to the center. So to put 80 turns on with the wire and stator I was experimenting with, I would plan to do the three layers like this I think:

28 ----> = 28
28 <---- = 56
14 ----> = 70
10 <---- = 80

If the turns on two adjacent arms used up all the room in the bottom of the "V" I would move the 3rd and 4th layers out a little ways, on one or both arms, to get some more room for the final turns.

I have not actually completed winding a full phase yet, and the counting is really an issue for me, it takes so much concentration. I think that if I try it again, I am going to use the 0.61"/15.5mm per turn to estimate the length of wire needed for one arm. That would be 0.61"/15.5mm x 80 turns = 48.8"/1240mm.

I would wind out that much wire and put a small temporary tape marker on the wire and spool it back up. Then I could wind with a cheerful abandon and not worry about counting turns. When the piece of tape showed up or got close I would measure and mark another phase and put the tape there.

Oops! I forgot something there! The 48" length will only wind one arm and I need one continuous piece of wire to wind four arms with. So I would wind out 4 x 48.8"/1240mm or 195.2"/4960mm of wire and put three small tape markers on it to mark the end of each 48.8" length so I would know I have to move on to the next arm.

I really think, on these high turn counts, that I would prefer to start with a known length of wire marked as four equal length segments over trying to count turns as I worked.

I'll take a couple of more photos on stripping and terminating the ends of the windings. There is a lot to be learned the hard way on that too...

Jack

Thanks! Great info.

Then I did a new test to day

Motor: hobbyking nx-4006-530kv
24n22p

Winding: 2 strands 0.15mm wire 30 turns per tooth.

7.3 ohms across 2 phases.

SBGC AlexMos test 2 (1 min 22 sec)

@ culexus: How is the torque on that?

Seems like you could 60 turns with 1 strand and hit 15 ohms, but that motor is a bugger to wind. But if you do it, I would be very interested in a comparison...

JussiH

I will try to wind a new one just with one strand and hit the 15 ohms, the torque is not bad, I only did the 24g hextronik before this one but there is allot more torque on this Nx-4006...

Is it better to hit the 15ohm than 7,3? This I don`t understand. what is the best?

Wow! That really looks good!

But 22N (slots or arms) 24P (poles or magnets) does not work according to the http://i.caendle.de/dev/test2/ calculator (slots has to be divisible by 3) so I am guessing that you have a 24N22P motor, right?

And the result for that looks really good. You have 264 cogging steps for finer movements and a very good winding factor.

Is the AaAabBbBCcCcaAaABbBbcCcC winding scheme shown in the image the one that you used?

Added note: I noticed that in the prop testing data at HK they identify the motor as a 24N22P motor. I also notice that they say it was wound with 22 turns, does that seem to be what you found? And was it terminated with a Wye bundle in the original wind?

Thanks for the help on the details.

Jack

Originally Posted by jackerbes

Wow! That really looks good!

But 22N (slots or arms) 24P (poles or magnets) does not work according to the http://i.caendle.de/dev/test2/ calculator (slots has to be divisible by 3) so I am guessing that you have a 24N22P motor, right?

And the result for that looks really good. You have 264 cogging steps for finer movements and a very good winding factor.

Is the AaAabBbBCcCcaAaABbBbcCcC winding scheme shown in the image the one that you used?

Jack

Thanx for correcting me It is 24N22P.. hehe..

Will try to wind it with 60 turns also to see the difference, but how is it really? is it better to aim for the 15 ohm or stay with the 7 ohm?

Well, I am also new to this, so I may be wrong....

But my basic understanding is, that the more turns you can get in there the more torque you will get from the same current. So in your case you should get more torque with 60 turns/1 strand. It seems like a good motor to try that on - the 4404 would take 200 turns of 1 strand 0.16, and I am not sure I have the patience for that....

Alex sent me some calculations that I am still trying to wrap my head around. As soon as I do, I will share it with you.

I'm pretty good on the motors but JussiH is the man with the answers for questions about which is better on the resistance.

I am a newbie too as far as winding gimbal motors.

@JussiH, (or anyone else that knows for that matter)

I've been browsing the AutoQuad pages and forums and am getting a little smarter, those are great resources for us new guys. One question I have that I can seem to find an answer for is if the AQ6 AutoQuad controller can be used on a tricopter. I see it does quads, octos, and hexs from the motor mixing table setup descriptions. But there is not any mention if it can be used on a tricopter or three arm copter. Will the AQ6 not work for tricopters?

I have been toying with building a tricopter just to ease into this a little and because I have a lot of the parts for that. I was going to use a KK board for that and probably still will but was wondering if the AQ6 is capable of being used on a tri.

Jack

Originally Posted by jackerbes

I'm pretty good on the motors but JussiH is the man with the answers for questions about which is better on the resistance.

I am a newbie too as far as winding gimbal motors.

@JussiH, (or anyone else that knows for that matter)

I've been browsing the AutoQuad pages and forums and am getting a little smarter, those are great resources for us new guys. One question I have that I can seem to find an answer for is if the AQ6 AutoQuad controller can be used on a tricopter. I see it does quads, octos, and hexs from the motor mixing table setup descriptions. But there is not any mention if it can be used on a tricopter or three arm copter. Will the AQ6 not work for tricopters?

I have been toying with building a tricopter just to ease into this a little and because I have a lot of the parts for that. I was going to use a KK board for that and probably still will but was wondering if the AQ6 is capable of being used on a tri.

Jack

Hi Jack

We dont support tricopters, but I think it should be possible to do so. Basically a question of the right mixing. Its on my (massively long) todo list to try it out.

I seem to remember that we discussed it in early development stages, but I cant really remember what we ended up on. But I dont see any reason it should not be possible to create a tri mixer for AQ. I could be wrong, though.

AQ was designed with advanced photo/video/inspection use in mind, so we did not focus on tricopters, since most guys into aerial photo and video tend to prefer the redundancy of 6 rotor (or more) crafts.

I dont mind answering questions on AQ here, but for the benefit of the community, I would suggest that you ask questions on our forum so that others may benefit from the discussion and answers.

forum.autoquad.org

As a note,IMHO tricopters are a lot of fun to fly, but they are not really any simpler to build than a quad. For funflying, the tricopter is a very good craft, but in my opinion they have no advantage for video and photos.

If you want to build a tricopter, there are other controllers out there that will be more "fun" to fly than the AQ - the AQ controller focuses on high stability and great position lock for photos, videos and inspection work. We dont have an "acro" mode, for that exact reason and it was not built with "funflying" in mind.

Also, as a first build the AQ may be way over the top, you would be in for a steep learning curve on calibration and PID tuning with that.

For a first build, I would suggest to get a Multiwii, KK 2.0 or a CC3D, since those will allow you to get to the fun part of flying a lot faster. None of them are as stable as AQ, but they are a lot more "fun" to fly and you would learn a lot about multirotor dynamics and tuning from those. AQ is an "expert" controller for people with experience and patience - its not recommended for beginners.

The basic of the 5-15 ohm resistance idea, is that the higher the resistance, the less current the motor will need to produce the same torque.

Remember that at the moment this is still in its infancy, and we will need to do some comparative tests on the same motors to work out what is best for a given task. That was why I asked culexus to try and wind the same motor with 1 strand. I will (when I feel up to it) try and wind a 4404 with 200 turns/1 strand of 0.16 and see if that gives more torque at the same power setting, but the idea of 200 turns * 12 is creeping me out a bit...

The 5-15 ohm should be considered a "ballpark" figure, but I have seen examples of much higher resistance motors that still worked well.

I will try to rewind it as soon as possible! If I get more than 60 turns I will do it.

Do you have a idea for testing the torque ? I will try to make a test and see if I can move a object or something just to see if the torque increases.

Originally Posted by culexus

Then I did a new test to day

Motor: hobbyking nx-4006-530kv
24n22p

Winding: 2 strands 0.15mm wire 30 turns per tooth.

7.3 ohms across 2 phases.

http://www.youtube.com/watch?v=5XmOCwwmI6Q

Looks real good!


Sent from my iPhone 5 using Tapatalk 2

Originally Posted by culexus

Thanx for correcting me It is 24N22P.. hehe..

Will try to wind it with 60 turns also to see the difference, but how is it really? is it better to aim for the 15 ohm or stay with the 7 ohm?

Hi Will this be 60 turns two strand? Or just 60T single?

Thanks,

Mike

Have this on order to see how it works.

http://www.rctimer.com/index.php?gOo...1&productname=