DYS1104 motor rewind for higher Kv - RC Groups
Shop our Airplanes Products Drone Products Sales
Thread Tools
Jun 21, 2017, 07:12 AM
Brisbane, Australia
Mini-HowTo

DYS1104 motor rewind for higher Kv


I'm going to attempt to rewind some DYS1104 5400Kv to around 9000 or 10000Kv for a friend to try. I decided to document it so that any others interested in rewinding these tiny little motors might find this useful to help get started in rewinding.

So why would we want say 9000Kv or 10000Kv - the reason is we are using very light 2.5" frames, and light 300mAh 2S batteries, and so far, Kv's of 7500Kv and 8000Kv have been performing very well, so we want to find the Kv upper limit for this setup where there is no point going any higher in Kv.

A bit of background on myself - I am a total novice rewinder having rewound about 7 sets of cheap quad motors from 2403 size down to 1105 size. I have rewound 2 sets of 1105/1106 motors and they are more difficult than larger motors due to the tiny size. So please keep this in mind if you follow this thread, that I am still learning as I go, and by no means good at this. I also encourage more experienced rewinders to correct me or help me if/when I veer off track.

Also, I am extremely thankful to all the motor rewinders who have shared their knowledge on rcgroups, without that, I probably could not have started rewinding at all.

I had these DYS1104 5400Kv motors on an old 3" frame, it had not flown a lot from memory and checking the frame and motors for dings and scratches confirms this. I put some props on it 2 days ago and flew it again and the motors sounded quite OK, so they were chosen for this rewind project.

Next I disconnected the motor wires from the esc's to be ready for the drill press which is used to measure the Kv. By knowing the (real) Kv, and counting the turns on each tooth when I unwind the stators, I can get an idea of how many turns are needed for the rewind to get to 9000-10000Kv.

The drill press spins at 892 rpm, I have measured it several times using a multimeter set to Hz and connected to 2 wires of a motor. From the frequency (Hz) and the number of magnets in the bell, the rpm can be determined. It is even more accurate if your multimeter can do min/max/average, the average is more accurate.

There is a lot more detail on motor Kv here. Basically, by knowing my drill press rpm is 892, I can abbreviate the Kv equation to Kv = 664/V where V is the AC volts generated by each winding of the motor when being spun by this drill press (note: 125mV = 0.125V). I have a 2mm aluminium tube in the chuck with some silicone tubing, the tubing goes over the motor shaft and spins it when the chuck is lowered into position over the motor shaft. Pic below of motors and drill press.

I wrote down the millivolts of each of the 3 phases of the 4 motors to try to get a full picture. I have found the turns count on each tooth to vary by 1 turn on some motors, and this makes it difficult to predict the new intended Kv. With these motors, the millivolts vary a bit from 124mV to 130mV (Kv varies from 5355Kv to 5108Kv). So here we have a variance of around 5% which is not too bad really, in my limited experience.

When the stators are unwound, I will note the turns on each tooth and see what I find. Disassembly will be the next step. Any questions, just fire away.
Last edited by Ian444; Jun 24, 2017 at 02:02 AM.
Sign up now
to remove ads between posts
Jun 21, 2017, 12:55 PM
homo ludens modellisticus
Ron van Sommeren's Avatar
Kv = 10,000rpm/volt and 2s? → 80,000rpm no load, and 50,000 with load.
Typo somewhere in your numbers?
If not a typo, why such a veeery high rpm?

Vriendelijke groeten Ron
Last edited by Ron van Sommeren; Jun 21, 2017 at 04:24 PM.
Jun 21, 2017, 04:50 PM
Brisbane, Australia
Hi Ron, no typo, 7500Kv motors in this size are already commercially available. One of my quads is running 1106 8000Kv on 2S, performance is great, motors and battery just getting a little bit warm. I think its possible due to the tiny scale and 2.5" (2535) props I am using. Others are also reporting very good performance with 7500Kv motors. Using lower Kv motors simply gives less performance.

Also, for well over a year I think, people have been flying the RotorX Atom 122 quad with 1105 6500Kv motors on 3S with the same 2.5" props. I'm hoping to find the "sweet spot" for 2S.

And thanks for all your work in this forum, keeping us informed, managing the sticky etc, much appreciated
Last edited by Ian444; Jun 24, 2017 at 01:56 AM.
Jun 22, 2017, 06:48 AM
Brisbane, Australia

Removing the e-clip.


The next step is to remove the e-clips/c-clips without losing them from flying away and never to be found again. Some people do the work inside a plastic bag to ensure the circlip can't escape. I've been using a piece of sticky tape and that has been working for me so far. If there is a bit of oil residue around the bearing it's worth cleaning it up so the sticky tape stays there. Be assured that if you don't use some sort of restraining mechanism, you will lose some clips, been there, done that. After putting the tape on, I get some small long-nose pliers and squeeze between the open end of the c-clip and the shaft, to move the c-clip just enough to get a needle (held in a pin vice) into a gap on the other side of the clip to pry it from the shaft. So the needle actually does the main work to pry the c-clip free. Once removed, I sticky-tape the c-clip to a piece of paper so it can't be lost. Then the bells are removed and the magnets cleaned with a cotton bud soaked in isopropyl alcohol. This is to remove any tiny magnetic particles from the magnets and generally clean the bell. If any magnetic particles remain, they can be removed with sticky-tape and tweezers, using the sticky side of the tape to cling to and pick up any stubborn particles. Once clean, I wrap up the bells and the c-clips and put them in a bag until I need them for re-assembly.

Just for info, the Racerstar 1104/1106 motors use a larger c-clip and it is a fair bit easier to remove/work with. The DYS c-clips I think are the smallest I have come across. They do come off fairly easily though with a needle in a pin vice. On the other hand, Racerstar 1104 windings are untidy in comparison. Pics:
Last edited by Ian444; Jun 24, 2017 at 03:26 AM.
Jun 22, 2017, 08:00 AM
Test Pilot
teracis's Avatar
Great write up, it will be interesting to see the performance of these!

Is this going to get close to the eRPM limit of the ESC's?
Jun 22, 2017, 09:43 AM
the people squad
schrodingers cat's Avatar
Fantastic detail, this will be a very useful resource for me when I give motor rewinding a go. Already a great tip with the sticky tape to catch the e clip
Jun 22, 2017, 03:54 PM
Jack
jackerbes's Avatar
If you want to test those to find the new Kv you can remove the outer layer of turns from them and repeat the test to find the new Kv. You could even remove a turn or two from the inner layer if necessary to find the Kv you want.

Then it is just a case of deciding the wire size that will give you a good fill for the turn count that gives you the Kv you want.

Those are 9N12P motors and should have an ABCABCABC winding on them.

In theory, halving the turn count will double the Kv. So I think you want to remove less than half of the turns and looks like the outer layer will do that.

Jack
Jun 22, 2017, 05:58 PM
homo ludens modellisticus
Ron van Sommeren's Avatar
Ian, such high rpm will give higher eddy current losses (proportional to rpm squared).
You could go from 12 to 6magnetpoles (same ABC winding diagram) to reduce eddy current losses. Could also be handy if you were to run into the eRPM limits of controllers.
Halving the number of magnetpoles will also roughly double Kv (and motor will reverse).

Vriendelijke groeten
Last edited by Ron van Sommeren; Jun 22, 2017 at 06:19 PM.
Jun 23, 2017, 03:18 AM
Brisbane, Australia
Quote:
Originally Posted by jackerbes
If you want to test those to find the new Kv you can remove the outer layer of turns from them and repeat the test to find the new Kv. You could even remove a turn or two from the inner layer if necessary to find the Kv you want.

Then it is just a case of deciding the wire size that will give you a good fill for the turn count that gives you the Kv you want.

Those are 9N12P motors and should have an ABCABCABC winding on them.

In theory, halving the turn count will double the Kv. So I think you want to remove less than half of the turns and looks like the outer layer will do that.

Jack
Thanks Jack, I have not tried this before but is it possible to wind just one phase (3 teeth) and put the motor in the drill press to measure Kv? I think it should work.

The motors have 9 turns per tooth so I'm thinking 5 turns wye terminated (wye is the original termination), and that should get me close to 9 or 10 thousand Kv.

Quote:
Originally Posted by Ron van Sommeren
Ian, such high rpm will give higher eddy current losses (proportional to rpm squared).
You could go from 12 to 6magnetpoles (same ABC winding diagram) to reduce eddy current losses. Could also be handy if you were to run into the eRPM limits of controllers.
Halving the number of magnetpoles will also roughly double Kv (and motor will reverse).

Vriendelijke groeten
Thanks for the info Ron. I had not thought about eddy current losses, I guess they will manifest as heat and less than expected power. I will keep this in mind for sure.

Below are pics of an 8500Kv motor and a DYS1104 10000Kv motor and although they were supposedly going to be released a month or so ago, they have not been released yet. But the Racerstar 1106 3800Kv motors I rewound to 8000Kv are performing really well with minimal temp rise and low battery stress (using a 300mAh 2S battery) so I will stick to Plan A for now, and see how it goes. If it does not turn out well, I now have another option, thanks.
Last edited by Ian444; Jun 24, 2017 at 01:57 AM.
Jun 23, 2017, 04:14 AM
Brisbane, Australia

Removing the stator.


On the motors I've been rewinding so far, it looks like the stator is glued to the bearing tube and base assembly with epoxy glue, probably a fast setting one like 5 minute Araldite or similar. The wires from the motor are cut as short as possible and the motor base is screwed to a piece of aluminium plate that I have for this purpose. Then the aluminium plate is heated (underneath the motor) until the upper end of the bearing tube is almost too hot to touch, this softens the epoxy. Most of the heat to the bearing tube can be transferred through the mounting screws and the aluminium plate so it's not necessary to actually heat the stator directly. If you heat the plate and put your finger on the upper bearing and bearing tube, you will feel the heat arriving up through the base. I use a gas flame from a gas soldering iron to do the heating, with the flame turned down as low as possible. It usually only takes about 15-30 seconds of heating with a very small gas flame. An alternative is to heat the whole assembly with a heat gun and heat it until almost too hot to touch.

Once the bearing tube is hot, I put 6" vise-grips with "soft jaws" on the stator and rotate it to break the epoxy bond, and try to work it up and off the bearing tube while rotating it.
Last edited by Ian444; Jul 08, 2017 at 09:52 PM.
Jun 23, 2017, 04:59 AM
Brisbane, Australia

The winding configuration.


This motor is 3 phase so has 3 individual windings each made from a continuous piece of wire wound around 3 teeth. So phase 1 would be wound around the 1st, 4th and 7th teeth, phase 2 would be on the 2nd, 5th and 8th teeth, and phase 3 on the 3rd, 6th and 9th teeth. All teeth are wound in the same direction, usually CW, when viewed from the end of each tooth. With the wye termination, the 3 wires that end up at teeth 7, 8 and 9 are soldered together and covered with a short piece of heatshrink. With 9 stator teeth, and 12 magnets, the configuration is known as 9N12P. In its stock form, this particular motor has 9 turns around each tooth using 0.25mm wire. It is wye terminated and wound according to the pic below, which I think has originated from this web page. The rewind of this motor will be exactly the same except there will be 5 turns per tooth to give 9/5 x current Kv.
Last edited by Ian444; Jun 24, 2017 at 02:05 AM.
Jun 23, 2017, 07:10 AM
Brisbane, Australia

Unwinding the stator.


Now that we know how this motor was wound, it makes it easier to unwind it. It seems a lot easier to unwind a motor than it is to use any other means to remove the copper.

Usually to unwind a motor you undo the last bit of winding first, which would be tooth 9 (the last tooth of the 3rd phase), and remember teeth 7, 8 and 9 are part of the wye termination/solder joint, so are easy to find. So one of these 3 wires will have been the last to be wound onto this motor. The solder joint of the wye termination is untwisted to get maximum length and then cut off to free up the 3 wires, and then you can see which wire is most likely to be the last one wound on the stator. Unwind the one you choose, and then it will lead to the previous tooth etc until that phase is totally removed. It takes a long time to explain it, but it's a lot easier when you are looking at it.

Sometimes, some epoxy gets on the wire, and holds it down preventing you from unwinding it. Gentle force can free it but too much strain will break the wire and you really want to avoid that as far as possible. If there is some epoxy holding the wire down, try to chip away some of the epoxy. If there are overlaying winds from an adjacent tooth trapping the wind you want to remove, try unwinding that other wire first, just enough to free up the wire you are working with.
Last edited by Ian444; Jun 24, 2017 at 03:47 AM.
Jun 23, 2017, 07:24 AM
the people squad
schrodingers cat's Avatar
You wonder if there is any "real" 1104 out there... Are all the 1105 stators you've measured 5mm tall too?
Last edited by schrodingers cat; Jun 23, 2017 at 07:34 AM.
Jun 23, 2017, 10:16 AM
Jack
jackerbes's Avatar
Quote:
Originally Posted by Ian444
Thanks Jack, I have not tried this before but is it possible to wind just one phase (3 teeth) and put the motor in the drill press to measure kv? I think it should work.

The motors have 9 turns per tooth so I'm thinking 5 turns wye terminated (wye is the original termination), and that should get me close to 9 or 10 thousand kv. ,snip>
It is possible, and done regularly. The advertised Kv will be the least accurate Kv as far as how it relates to future Kv predictions. Your use of the drill press method will give you the most accurate Kv, a raw Kv (full throttle, no load, RPM / Voltage) will be very close to the same value and is much quicker.

In the two images I salted Turn Calculator 5 with the advertised 5400 Kv for the first image and then changed that to 5278 as if I had measured a raw Kv for the second image. You can look at the values generated for Delta and Wye terminated LRK, Half Parallel dLRK, and dLRK there and choose the one you want to try. I generally will use the one that has the lower or lowest turn count so as to let me use the largest wire that will let me attain the turns needed.

You are working with 9N motors so you choice really comes down to only the ABC wind and changing to either Delta or Wye terminations. With 12N motors all of the winds mentioned are candidates. On 9N changing the magnet count from 12 to 10 (9N12P or 9N10P) will change the winding, you can look at images of that on Skylar's wonderful Homebuilt motors wind image calculator page by changing the poles count from 12 to 10. It changes the winding scheme from AaABbBCcC for 10P to ABCABCABC for 12P.

I use the free Open Office Calc spreadsheet to run the turn calculators and save them as Excel 97-2003 *.xls format as that is the format most commonly found here. There are other and smarter Turn Calculators too, TC5 is a good one for discussions though.

Jack
Last edited by jackerbes; Jun 23, 2017 at 10:29 AM.
Jun 23, 2017, 10:19 AM
homo ludens modellisticus
Ron van Sommeren's Avatar
Quote:
Originally Posted by Ian444
... this particular motor has 9 turns around each tooth using 0.25mm wire. It is wye terminated and wound according to the pic below, which I think has originated from this web page. The rewind of this motor will be exactly the same except there will be 5 turns per tooth to give 9/5 x current Kv.
Because:
Kv_new = Kv_old #winds_old / #winds_new
(for the same phase hookup)

Quote:
Originally Posted by Ian444
... is it possible to wind just one phase (3 teeth) and put the motor in the drill press to measure Kv? ...
Yes, you'd determine Kv_delta that way.
Kv_wye = Kv_delta / √3

Quote:
Originally Posted by Ian444
... I had not thought about eddy current losses, I guess they will manifest as heat and less than expected power. ...
Yes and yes.

Quote:
Originally Posted by Ian444
... usually CW ...
In fact, CW or CCW does not matter, that merely indicates opposite winding, all are/must be wound in the same direction, in a 9N/12magnet (or 6magnet) motor.

Quote:
Originally Posted by Ian444
... With the wye termination, the 3 wires that end up at teeth 7, 8 and 9 are soldered together and covered with a short piece of heatshrink. ...
Starts and ends of phases matter a great deal when hooking up the phases in star or delta. Get it wrong and one ends up with a dud.

Prettig weekend Ron
Last edited by Ron van Sommeren; Jun 28, 2017 at 02:45 PM.


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Motor Rewind for Low Kv rivers Electric Motor Design and Construction 24 Aug 25, 2016 12:35 PM
Discussion Hacker 20-22L rewind for higher Kv billyjones454 Electric Motor Design and Construction 11 Feb 03, 2012 08:48 PM
Discussion Rewinding a motor for higher KV Eddy Electric Motor Design and Construction 32 Jun 22, 2011 05:19 PM
Discussion Need Advice Rewinding BL Motor To Get Higher KV phupper Electric Motor Design and Construction 6 Dec 10, 2008 12:38 PM
Discussion Rewinding an axi motor for higher KV? Eddy Electric Motor Design and Construction 10 May 05, 2007 12:30 PM