Thread Tools
Aug 29, 2012, 05:56 PM
low tech high tech
vtdiy's Avatar

Making a CNC Milling Spindle from an R/C Motor

I've got a CNC router and am building a quirky sort of CNC turn mill from scratch. The router fits a Bosch colt router, but lately I've been thinking about using an R/C electric motor as a more accurate milling spindle for both machines.

This has been done before, but I thought it might be interesting to put the thread here because of the possibility of rewinding one or both to better suit them for the purpose. I don't know really -- but I'm sure there will be some great input for this project here.

I've been watching the sales on Leader Hobby and managed to pick up $15 and $17 motors of what seem like useful sizes for this -- and EMP N5055-580 and an N5055-400. (580 and 400 kv)I haven't taken them apart yet, but will soon.

I also purchased an 8mm ER11 spindle collet holder for $17 and a set of ER-11 collets from 1mm to 6mm. That was kind of a splurge but still reasonably priced, imported from China. I probably could have gotten by with a 6mm and 3.5 mm.

Anyway, the idea is to replace the 8mm motor spindle with the 8mm collet holder, and make a housing to support some bigger bearings.

The questions I have are more rewind oriented than mechanics.

First question, I have no clear idea what kV would be a good choice for this. I guess it depends on the power supply and the spindle speed range I need.

I would like to mill everything from foam to steel, if possible, so speeds and feeds of are going to be a really wide range.

Thinking about power supplies, it's easier to increase voltage than current in a line operated supply. Typically 10 amps at 36 volts DC are fairly easy to come by. And higher voltage may be possible. In fact I have the makings for a supply already. So that's say 350 watts.

Here's a picture of the parts I have so far.
Sign up now
to remove ads between posts
Aug 29, 2012, 07:57 PM
low tech high tech
vtdiy's Avatar
I guess the way to approach this is first decide what the maximum RPM I want to have is, then determine the kV from that and the running voltage.

Ideally I guess I'd want a high rpm for milling wood. Typically wood routers run at 15-25K rpm depending on cutter size. Lets say a 6mm cutter, so 18k - 20k RPM.

But I do wonder what the maximum safe RPM for an outrunner motor of this size is. Whatever that is would be the limiting speed and that would determine the target RPM, and therefore the target kV.

Question: Is 20,000 RPM way too high for a typical outrunner motor of this size (case diameter 50 mm -- motor shaft 8 mm) running without a prop?


If 20,000 RPM is permissible, then 20,00 RPM / 36 volts = 555 kV, which actually is pretty close to what these motors happen to be already.
Last edited by vtdiy; Aug 29, 2012 at 08:09 PM.
Aug 29, 2012, 08:51 PM
low tech high tech
vtdiy's Avatar
Here's the 580 kV motor opened up.

The stator diameter is 40 mm, stator tooth length is 25 mm.

12 stator teeth and 14 magnets.

Epoxy is a little sloppy on the magnets but none on the faces and the spacing between magnets looks even, by eye.

I didn't measure the gap between teeth and magnets.

I can measure the static resistance if that's of interest.

I see only three leads and no soldered and insulated 4th bundle, so I assume that means it's delta connected?

I don't have a caliper with me, measuring with a ruler I make the bearings out to be 17 mm OD, and of course 8 mm ID. Don't know what the depth is. The shields are extremely fragile -- a foil basically, as they were dented with light pressure from the circlip pliers posts.

(Had a hard time getting the circlip off -- it's in a recess in the motor face. Even my circlip pliers didn't have pins close enough to fit -- I finally bent the pins together to get it to engage. Also there was a burr on the circlip groove I had to file down before I could get the stator out)
Last edited by vtdiy; Aug 29, 2012 at 09:09 PM.
Aug 29, 2012, 09:11 PM
Registered User
erm Use/fit a servo tester/controller ?
hooked to an ESC .. subsequently motor RPM range is variable 'as required'.
You then get to use your Flying batteries too.
Still think this is for CNC forum :-)
Aug 29, 2012, 09:20 PM
low tech high tech
vtdiy's Avatar
Yes I know about controlling the speed, already have servo tester and ESC for that. Sorry, won't be using flying batteries.

This is about specs, winding and internal construction, calculation, and modification of R/C motors. If I'm boring anyone with that kind of thing, apologies.

Anyway looked at EMP advertised specs and that shows 4-8 cell rating for the motor. Working backwards, 8 cells is roughly 30 volts, and 30 times 580 kV is 17,400 RPM. That's no load. But seems like the motor should be able to turn that without the rotor coming apart.
Last edited by vtdiy; Aug 29, 2012 at 09:38 PM.
Aug 29, 2012, 10:52 PM
low tech high tech
vtdiy's Avatar
The kind of slightly disappointing thing about it is, looks like the 580 kV motor is just about exactly right as-is, without rewinding.

I was wondering what kV would be appropriate for this kind of project. Didn't know how to guess at it before I started thinking about it here. As luck would have it, it already is a good number.

The 400 kv motor would probably be better re-wound if wood was a material to be milled. Or if a higher voltage supply was available. For other materials it seems fine.
Aug 29, 2012, 11:09 PM
Wake up, feel pulse, be happy!
Piece's Avatar
The motor should be ok, but there are a few things to watch for.

First, on these cheaper motors the glue job on the magnets can be spotty. Just make triple-sure that all the magnets are glued in securely before cranking it up to like 18k. You just don't want them shifting or realigning while the motor's running, obviously.

Second, bearings. Using the thing for milling will probably put some side loads on the bearings. Whether it'll be more or less than a prop's gyroscopic forces in flight (e.g. while pitching/yawing), I don't know. Just be mindful of it, and maybe measure them in advance so you can replace them when/if necessary. Some good high-quality bearings may not exactly be cheap (probably cost more than the motors themselves ), but if you're planning on doing anything more than just fooling around with this thing then they're most likely worth it.

I've thought of this sort of thing before, but for me it was always related to the fact that Dremel bits all seem to have 1/8" shanks and could be dropped right into many of my parkflyer motors.
Aug 30, 2012, 06:27 AM
low tech high tech
vtdiy's Avatar
Thanks C4H10, I'll look at the magnets more carefully.

I forgot to mention that I will be adding bearings external to the current motor housing. There will be a bolt on spindle housing I will make up -- the new shaft is quite a bit longer than the old one. The 2 bearings in the spindle housing will be much larger ABEC-7 608Z -- I already have those. They are roller skate bearings. So side loads will be taken there.

It might still be a good idea to to replace the motor's bearings so I would like to find a source for those. I don't know the thickness yet, though. Just the ID and OD.

A couple of points open to suggestions:

Current motor end plate is press fit on the shaft, and I could remove it and press fit in the new collet shaft. But that means all bearing sets would need to be mounted first because they won't be able to slide over the collet end. It would be nicer if the shaft were removable without pressing in or out.

The other concern is how I will limit current to what the power supply can put out as a maximum. A battery is different than a line power supply, and usually in a plane you pick out a battery sized to deliver more than the normal load of the prop/motor combination by a good safety margin. A miller spindle can stall (not so likely in a plane motor) so the max current would be the stall current.
Aug 30, 2012, 06:28 AM
jackerbes's Avatar
The best way to spot the wind type you have now is to look for for the long transit runs that go 180 degrees across the motor, only dLRK will have those. That and the fact that there will be a short transit from one arm to the next in the side by side pair of arms that make up the quarter phases.

An ABC wind would have three starts side by side and and each winding would wind an arm, skip two arms, then wind another.

I did a lot of machine work in my years as a machinist, on lathes and milling machines. There is a need for a wide range of spindle speeds to handle cutting in different materials. For milling, it is the number of teeth on the cutter and the feed rate/chip load that decides what will work.

I had a light milling machine (it was actually a Linley jig boring machine) that had a Dayton 3/4 HP PM 90V DC motor on it and that was very versatile but you had to be satisfied to let the cutter and material decide how quickly the work could be done. It was used for one time tasks and accurate work, it was not a machine for mass production.

The 90V DC motor on the Linley was rated at 9A and and it could pull that much current and cut material down to a very low RPM, even as low as 50 RPM or so.

Last edited by jackerbes; Aug 30, 2012 at 01:20 PM.
Aug 30, 2012, 08:48 AM
low tech high tech
vtdiy's Avatar
Thanks Jack. I will look at the windings more carefully. They are hard to see with the end plate on. I won't be pulling them off unless there's a reason to re-wind. Maybe someone will give me a reason to do that!

This motor is actually rated at 1580 watts. Your motor was 90v x 9 amps = 810 watts, so there's potential for reasonably precise non-production spindle, as you say, at a more conservative wattage.

I do have a 70 amp ESC for this project. But my real limitation is the power supply. I'm certainly open to suggestions there. My assumption (maybe a wrong one) was that it's easier to get or make a high voltage supply than a high amp one. High current transformers are hard to find and expensive. I have a couple of 10 amp xformers, some big filter caps and rectifier bridges thast could handle it.

I can definitely make a 10 amp 36 volt DC supply with parts I already have on-hand. That's only 360 watts, but for hobbyist/modeling use with cutters 1/8" to 1/4" -- often in foam or wood for fine carving, that might be fine. Occasional metal would mean slow speeds and feeds, but for limited use I think it would be okay for me.

It certainly would be nice to up the wattage some to take better advantage of the motor.

I would not want to use Lipo flight batteries in a CNC miller for many reasons. I suppose you could use a marine deep cycle lead acid battery ( which I happen to have) with more capacity and greater safety, but I really don't like the idea of using batteries and re-charging, etc on a machine that may be left unattended for even short periods of time.

So a line operated high current low cost supply would be the best solution -- any suggestions or alternatives?
Aug 30, 2012, 10:01 AM
Dieselized User
gkamysz's Avatar
Here's my story.

I suggest the power supply be capable of twice the anticipated current draw. To get a real good idea of motor loads you should calculate them based on the cutting specs you intend to use. Every once in a while you'll get that odd tool path that will throw a full width pass into your code. If the powersupply cuts out, you'll have at minimum a broken tool. The spindle should be able to handle a 150% of rated load indefinitely, and at least 200% for a short period. By rated load I mean the expected cutting loads. You're 2HP motor will never see 2HP while cutting. The power supply voltage doesn't matter much if you can change the motor Kv. Just give yourself plenty of headroom. If there is some way to get variable supply voltage that would be ideal if you need a very wide range of RPM. Cheap RC ESC will not allow running a large diameter cutter at lower speed with enough torque at low ESC PWM duty cycle. I'd think about converted server supplies that can be run in series which you could connect 1,2,or 3 in series to get closer to the RPM you need.

If you are doing any sort of work with cutting tools smaller than 1/8" spindle runout will be very important. Don't settle for anything worse than .0005" TIR. Even that is a lot for <1/8" 4 flute endmills at high speeds and feeds. Chances are pretty good the collet taper will not be true. The first shaft I had I got lucky with and was dead nuts. after bending it, the spare I had actually had more runout out of the box than the bent one! I had to true the collet taper after installing it in the spindle.

Next is vibration. The spindle must be balanced well in order to perform. Surface finish will depend on it. Machine life will also be effected by vibration. 20k on a spindle that hasn't been dynamically balanced to G3.5 or better is a waste of time, IMO. It won't feel bad when held in your hand, but mounted to the machine it will sound like it's going to shake the machine to pieces.

Bearings. I went high end on the bearings. The bearings must be preloaded to survive high speed and eliminate shaft end play.

Chips at cutting speeds of 15k are small and go everywhere. They will get into your motor if left open like that. Obviously, that motor relies on air cooling through the motor which is going to be a problem.

I bent my 20mm spindle shaft by breaking a 1/4" carbide endmill. So don't expect much from that 8mm ER11 shaft. Wood is probably fine, aluminum with 1/8" end mills maybe, steel I don't know.

My project was intended to allow cutting steel with 1/4" endmills at up to 15kRPM. It will do that, but leaves little room for error. It should run to 30k, but I haven't tried it that fast yet. I set some high standards for my spindle and am happy with the results, but for the bent shaft. You might be happy with something of much simpler construction as long as you know what you're giving up. My site needs a bit of updating with the crash info and water cooling which I added.

Aug 30, 2012, 11:07 AM
Scratch builder
Will a sensorless motor work in this application without losing timing?
Aug 30, 2012, 11:14 AM
Dieselized User
gkamysz's Avatar
Yes, I've been using a Castle Creations ESC so far. The sensors in my design are for the industrial type control I haven't added yet. I've only run it at full throttle on a variable bench supply.

Aug 30, 2012, 12:48 PM
low tech high tech
vtdiy's Avatar
Thanks Greg! That is a carload of important information. Also a great website.

I do worry about stalling/crashing/power supply problems. Seems clear if the cutter stops, it is crashing, period unless there is some kind of sensor emergency stop for low spindle RPM -- is that something you have now on yours?

Also thinking I should maybe reduce my expectations. I would say the essential material would be foam profiling for models. And as few flutes as I can get, given a long mill length. Speeds for foam actually shouldn't be way high because of the low melt temp of the foam. Feeds will be high to get chip size up. Cutting depth should be capable of at least 2" I hope,b ut it gets hard to find cutters above that length. I could go to bigger cutters than 1/8" to get speed down -- 1/8 is just nice because the waste is less. But waste isn't the biggest concern. I don't see a need for very small pocketing, or even drilling with foam. So cutters don't have to be tiny if that helps make a spindle easier.

If it cuts foam well, next in preference would be wood. But that one has pretty high demands, I think. I'd like to do some fine carving work on small workpieces. So small tools for wood.

Aluminum isn't essential. Steel definitely not. Nice if possible, but not requirements. I do have non-CNC machines for regular metal machining. I'm not interested in production at all.

Also I didn't realize that you could series server power supplies -- aren't they regulated? Apologies if I'm not understanding that properly.

Finally I think the difference between an engine building machine like yours and essentially a soft material carving machine, like what I'm envisioning is the acceptable tolerances for the finished parts. I'll be cuttting out models -- more of a roughing operation, not bearing surfaces. So I hope the requirements are do-able. I do want to balance and reduce runout, but to prolong spindle life more than produce close tolerances on foam or wood parts.
Last edited by vtdiy; Aug 30, 2012 at 01:16 PM.
Aug 30, 2012, 01:30 PM
jackerbes's Avatar
"..This motor is actually rated at 1580 watts. Your motor was 90v x 9 amps = 810 watts, so there's potential for reasonably precise non-production spindle, as you say, at a more conservative wattage..."

I added some wind images there, I forgot them before. That was a belt driven spindle and the 90V DC motor was an industrial type motor, a long permanent magnet inrunner I think we would call it. I knew less about motors then, but that worked well. It had small (8" square or so) power supply that was 120V AC in and 90V DC out and it had a simple knob for a speed control and a switch for reversing it. From what I know now, I think that was a switching type PWM power supply.


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Help! i'm looking for a tiny video cam for an r/c jet iflyrc1012 Aerial Photography 1 Jun 24, 2012 05:00 PM
Wanted Mills 1.3 non r/c carb.... Trade for r/c carb stegla Aircraft - Fuel - Engines and Accessories (FS/W) 0 Jun 04, 2011 03:30 AM
Wanted Wanted: Radio/motor/battery setup from an el-cheap-O "RTF" R/C plane. Cheesehead Aircraft - General - Miscellaneous (FS/W) 1 Aug 01, 2006 11:05 PM
Can I use an r/c car battery for an r/c plane?? jacksonjrs Parkflyers 7 Jan 28, 2004 10:25 PM