Using RC Brushless motors as spindle

[Jonathan]

Following on from the ideas in 'blackburn mark's thread 'Fingers crossed it'll cut ally' I'm going to document my spindle separately as that thread is getting a bit long...

This is the motor I'm using:

http://www.hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=5142

It will do 6000rpm on 48v, and more power than I'll ever need. I would have bought the 180kv version to get a bit faster if I had not been intending to make an electric bike with it at first!

I've bought an ER16 150mm long 12mm diameter collet chuck along with a set of collets for it which has finally arrived after well over a month. The general idea is to simply replace the existing 12mm shaft with the chuck and put a single double row angular contact bearing at the front for good measure.

Below is a rough drawing of the motor - I'm not exactly sure of the positions of the bearings inside it. There's three (might be 4) 6801 bearings and one large thin section 6809 (61809 on skf site) bearing near the front of the motor:

http://jonathan.blissett.me.uk/images/forum/Brushlessmotorspindle1.png

(At risk of sounding patronising, remember this is an outrunner motor - so the casing etc rotates.)

I've drawn in the angular contact bearing mounted in a bit of aluminium bar which will bolt nicely onto the existing M5 holes in the motor. The other bearings are left to their own devices.

My problem at this stage is the collet chuck has presumably been ground to 12.000mm, or pretty close so the bearings will just require a press fit. So my problem is how do I fit the shaft through all those little bearings without damaging them. I can't really remove them from the motor since they're stuck in with some sort of adhesive. If all else fails I can remove the part from within the stator that holds those three bearings entirely, and make a new version with better bearings.
The angular contact bearing isn't an issue - I can press that onto the chuck first, then the housing. Should get it nice and accurate...

There's a convinient M8 threaded hole in the end of the collet chuck which I could use along with a thrust bearing to preload the angular contact bearing against the thin section bearing. Having looked at the load ratings for those bearings I think that would be ok. Maybe I should just get another angular contact bearing instead of a thrust bearing, but that would stop me being able to disassemble the thing.

So er...any advice on fittings these bearings or the design would be most welcome! I'll machine it tonight if we come to any reasonable conclusion.

[Jonathan]

Just noticed a significant error in my drawing - the bore through the red bit is a constant diameter, so the bearings would come out from either end if it wasn't for the adhesive. I could try warming up the stator/bearings together to loosen the glue... I don't want to break anything though as this was quite expensive! If the right glue is used then this shouldn't work, I think, because otherwise they'd come loose during normal operation.

I wonder if heating the motor and freezing the collet chuck might make enough clearance to assemble it?

[blackburn mark]

personaly (because im lazy) i would mask of the section where the A/C bearing is going and wet n dry the rest down on the lathe to a more convenient diameter (im assuming your collet shaft is your cheapest part)
if thats a double row A/C bearing you wont need any pre-load on it

i thought you were going to make your bearing housing 43mm ?

[Jonathan]

personaly (because im lazy) i would mask of the section where the A/C bearing is going and wet n dry the rest down on the lathe to a more convenient diameter (im assuming your collet shaft is your cheapest part)
if thats a double row A/C bearing you wont need any pre-load on it

i thought you were going to make your bearing housing 43mm ?

It did occur to me to do what you said with the wet and dry, the problem is that will surely increase runout since all the other bearings will locate on an uneven surface.
It is double row, but I think it needs to be fixed onto the shaft somehow to stop it creeping.

Yes I was going to make it 43mm, to fit with Kress spindle - however I'm now going to buy Chinese 80mm spindle instead so etc. I may make it less anyway as annoyingly the smallest aluminum bar I have in stock is 3".

I've been reading up on heating bearings to make them fit. I think it should work for the angular contact bearing but I'm not sure about the others since they're housed in aluminum with laminated steel around that which will all expand at different rates.

[phill05]

mount the 12mm chuck in the lathe and skim a tad off between the bearing seats this will aid pushing through, freeze the part no need to heat the motor, should go in a treat have used this method myself.

Phill

[blackburn mark]

It is double row

good move in my opinion :) my double row was a press fit i havn't used locktite on it yet, i did intend to when id tried it and all looked well, its done some hard work so far and iv had no issues with it, i do get a kind of harmonic ripple in the finish when i use a fly cutter in aluminium but it performs way above my exspectations and the ripple may be being caused elswhere ? (gantry sides maybe)


80mm spindle
ah, with you now :)

i'd be shit scared of getting the shaft stuck only half way home using heat... that shaft will soak the heat out of those tiny inner bearing races in no time

i would think that runout issues of bearing misalighnment (im pretty sure you will get some no matter how hard you try) would be massive compared to any runout you get with the wet n dry inaccuracies
the advantage in my opinion to the microscopic missalighnment you will "probably" get is that it will pre load all your bearings to some extent and stop any of them free wheeling and skidding
However........ im no exspert and only guesstimating with this stuff :)



skim a tad off between the bearing seats

nice

[Jonathan]

mount the 12mm chuck in the lathe and skim a tad off between the bearing seats this will aid pushing through, freeze the part no need to heat the motor, should go in a treat have used this method myself.


That means skimming a tad off ('bearing relief' is the term I think) in three places. I may well do that. I'll need to accurately measure the bearing positions. I did it on the shaft for the tapered roller bearings on my wind turbine and that was effective. Though that time I made the shaft/housing from scratch...machining 80mm steel bar down to 25.4 took a while!

If the bearing did get stuck half way I can always use my big milling vice as a press. I'll go and machine a tube thingy to push the angular contact bearing on so I don't damage it.

The collet chuck is in the freezer now.

[Jonathan]

I put the motor back together, just to check it still works after splitting the 3 phases - it did. So I dismantled it, soldered connecters/heatshrink on, put it back together and it didn't work. Just twitched...

Found that two of the coils were shorted together :redface:
Dismantled it again, jiggled the wires a bit and the short went, so I put it back together and it now works fine. I think I'll spray it with some insulating stuff I've got for PCBs.

Anyway, now the big bearing rattles...the motor was silent before so I'll buy a new bearing. I'm not sure if I should get metal shield like the original, or rubber sealed due to all the dust. The rubber sealed bearings are only rated for 6500rpm though - bit close?

http://cgi.ebay.co.uk/BEARING-6809-2RS-61809-2RS-45MM-X-58MM-X-7MM-THIN-WALL-/200511636092?pt=LH_DefaultDomain_3&hash=item2eaf6cc27c

Good news - used emery paper to just to take a little off the very and and now the bearings inside the motor fit, requires a fair bit of force but not a press fit. Angular contact bearing is still a press fit.

Started machining the housing for angular contact bearing - slow progress on my little lathe!

[blackburn mark]

jiggled the wires a bit

i love it when that works :)

id be tempted to run it for a while, maybe the noisy bearing is a bit of crap thats snook into one during the trauma of the surgery you have done on it

[Jonathan]

i love it when that works :)

id be tempted to run it for a while, maybe the noisy bearing is a bit of crap thats snook into one during the trauma of the surgery you have done on it

I was thinking I'd have to rewind the motor. If I had had to I would have increased the kv.

I ran it for about 3 mins on 12.3v and the sound didn't change. It consumes 40w at that voltage with no load.

[Jonathan]

It's alive!

Heated the angular contact bearing in oven, cooled the shaft and it dropped straight on.
Bit of a fiddle to get the rest together, but it went ok.

I've run it up to 3100rpm (24v) and it sounds fine. Slight noise from the bearing but I don't think that's significant.

I also measured the runout and it's less than 0.01mm.

Will post some more pics in a minute.

3639
3642
3646
3643
3647
3644

[blackburn mark]

It's alive!

Heated the angular contact bearing in oven, cooled the shaft and it dropped straight on.
Bit of a fiddle to get the rest together, but it went ok.

I've run it up to 3100rpm (24v) and it sounds fine. Slight noise from the bearing but I don't think that's significant.

I also measured the runout and it's less than 0.01mm.

Will post some pics in a minute.

Nice one jonathan !!!! iv waited a long time to see this beast
it should have enough power to make a walnut wip out of your Y axis :)

[Jonathan]

Nice one jonathan !!!! iv waited a long time to see this beast
it should have enough power to make a walnut wip out of your Y axis :)

That makes two of us.
Make a walnut what?! But yeah, running it off two 12v 92Ah batteries for the moment. Not got a power supply which would supply anything like enough for this. Idling at 3100rpm it consumes 94w. 45w ish at 1500 rpm... that implies almost 200w just to spin the thing at full speed let alone cutting!

[blackburn mark]

looks great, i love how simple these thing are :)



Make a walnut what?!
3640
walnut whip ! .... shit joke.... sorry :)



94w

owch, but as M250cnc phil said to me not to long ago"give your bearing a chance to run in at low (ish) speed before you start flapping about heat" or "drag" in your case

im getting on fine using 350w 12v on my 1700w motor, i dont think you will need to throw 6000w at yours if your using it on your router

[Jonathan]

owch, but as M250cnc phil said to me not to long ago"give your bearing a chance to run in at low (ish) speed before you start flapping about heat" or "drag" in your case

im getting on fine using 350w 12v on my 1700w motor, i dont think you will need to throw 6000w at yours if your using it on your router

I did the mount today, not the neatest of things I've made with the rotary table but it works:


3654
3653
Spinning...nice effect:
3655

I ran the motor in for about 5 mins on 12v, soon got bored of that (I'll leave it running properly tomorrow). So I tried carefully grasping the motor body with some kitchen roll to protect my fingers - couldn't stall it, not even close :smile:. It merely drew 550w and about 53 amps. I'll have to think of a safer way of testing the power at higher voltages.

I think I'll stick with the batteries for now unless anyone can point me to a 50v PSU that can deliver a LOT of current. Maybe a few of those switching PSUs people have been using for steppers on eBay...

[M250cnc]

I did the mount today, not the neatest of things I've made with the rotary table but it works:

(Picture tomorrow)

I was under the impression your mill was under cnc control why the RT ?


I ran the motor in for about 5 mins on 12v, soon got bored of that (I'll leave it running properly tomorrow).

Its a temperature thing if it doesn't get hot just keep increasing the speed in increments.


It merely drew 550w and about 53 amps. I'll have to think of a safer way of testing the power at higher voltages.

I think I'll stick with the batteries for now unless anyone can point me to a 50v PSU that can deliver a LOT of current. Maybe a few of those switching PSUs people have been using for steppers on eBay...

Sounds like mum & dad are gonna have a big leccy bill coming. :eek:

Phil

[Jonathan]

I was under the impression your mill was under cnc control why the RT ?

I took the steppers off the milling machine and put them on the router. Didn't want to put them back just for one part. Rotary table was also the most accurate way I could make the hole for the spindle, not having a boring head.




Its a temperature thing if it doesn't get hot just keep increasing the speed in increments.
That's good, the bearing didn't get hot at all. ESC warmed up a little but that's all. I'll leave it running on 3100rpm for a while today and see what happens.


Sounds like mum & dad are gonna have a big leccy bill coming. :eek:

Hmm, didn't think of that - and nor have they :naughty:
The batteries I'm using to power it were going to be charged from my wind turbine. If only I had space to mount it...

I measured the runout again now it's mounted properly. I put a 10mm cutter in and measured 0.03mm about 10mm from the collet. If I put the DTI on the collet chuck itself it reads 0.01mm.

[blackburn mark]

I ran the motor in for about 5 mins on 12v, soon got bored of that

i wouldnt worry about it jonathan, just use it as you would and keep tabs on how hot the bearing housing gets, i still check mine now and again as im using them, its ony my high speed spindle that shows any signs of struggling and thats only if i push it to 30000/40000 rpm

phils point is confirmed as all three spindles produce less bearing heat as time goes by

[M250cnc]

I took the steppers off the milling machine and put them on the router. Didn't want to put them back just for one part. Rotary table was also the most accurate way I could make the hole for the spindle, not having a boring head.

4 jaw chuck on the lathe ?


I measured the runout again now it's mounted properly. I put a 10mm cutter in and measured 0.03mm about 10mm from the collet. If I put the DTI on the collet chuck itself it reads 0.01mm.

Are you measuring the outside or the inside of the chuck for runout ?

You should be measuring the inside and i would measure at the extremes of the taper to see if you have different amounts of runout at the two points.

Phil

[Jonathan]

4 jaw chuck on the lathe ?
Are you measuring the outside or the inside of the chuck for runout ?

You should be measuring the inside and i would measure at the extremes of the taper to see if you have different amounts of runout at the two points.


It wouldn't fit easily on the lathe - it's only a C3 mini lathe with 4" 4 jaw chuck. I probably could have got it to fit, but it wasn't worth it since I needed the rotary table to do the semicircle on the outside anyway. Wish I had a lathe like yours...made so many things on the same one at school.

I used the lathe with 3 jaw chuck for a few of the bearing mounts for the screws - carefully designed the shape to fit the chuck easily.

I measured the runout inside. I'll try again like you said.

Just tried cutting with it. 7mm two flute cutter and some scrap aluminium. Using 0.5mm depth/width of cut it was happy. Sometimes 1mm, machine is a bit too wobbly to do more at the moment. I'll post pictures in a minute to save a thousand words...

3659
3656
3660
3657
3658

The motor drew 30w extra whilst cutting, 100w idle.

[Jonathan]

i wouldnt worry about it jonathan, just use it as you would and keep tabs on how hot the bearing housing gets, i still check mine now and again as im using them, its ony my high speed spindle that shows any signs of struggling and thats only if i push it to 30000/40000 rpm

phils point is confirmed as all three spindles produce less bearing heat as time goes by

I ran it at 3100rpm for 20 mins. The angular contact bearing was very slightly warm - helped no doubt by all the aluminum around it. The stator got a little warmer, still nothing to worry about.

Managed to get 1.5kw out of the motor without it complaining.

I'll have to buy one of these soon to run it at 6200rpm:

http://www.hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=10944

or maybe

http://www.hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=4691

Either way it's a lot of money.

[M250cnc]

Jonathan

Use a fine oilstone to de burr the slits in the collets as they can have tiny burrs which can be a cause of runout.

When checking the runout it the collet mark the high or low point so you can see that they are in the same place vertically when you measure both positions.

Phil

[M250cnc]

Jonathan i have just seen the pictures of a sample of the cut produced by your machine with the spindle.

All i will say is you have balls of steel.

Phil

[Jonathan]

Jonathan i have just seen the pictures of a sample of the cut produced by your machine with the spindle.

All i will say is you have balls of steel.


What did I do wrong :redface: this time?

Part was securely mounted...it didn't move. I raised it up to stop the Z axis bending too much and took light cuts.


Thanks for the tip about deburring the collets - I'll try that.

[M250cnc]

Sent you a PM

Phil

Sent from my HTC Desire using Tapatalk

[discoball]

Hello
sorry for my bad english but you'v understood:i am french
i was very interested by your post
and am planning to put one BLDC motor on my home made CNC
can i ask you what kind of alim (not batteries)are you using for these motors and controller?
something like 24 or 48v and 40amp
i would be able to make one if i had a plan
thank you for your reply
Discoball

[Jonathan]

Currently I am using batteries, since I don't have a big enough transformer or power supply to run that motor.

One option is to get a toroidial transformer, such as one of these:

http://www.rapidonline.com/Electrical-Power/Transformers/Toroidal/Toroidal-transformer/82719

Alternatively it may be possible to put multiple cheap computer ATX power supplies in series to get the desired voltage.

[discoball]

thank you
for your fast reply Jonathan

i'll see that !!
I have not yet recieved my chucks

Thank you
Discoball

[templecorran]

Bravo!
I just started doing something similar.
Only I chose a ER11 C8 L100 spindle. When I got the bearings, I discovered why the spindle was so cheap! The shaft is not parallel. The spindle falls into the bearing with no effort at all. 8.00mm at collet end to 7.93mm over the length.
These spindles must be quality rejects. Thank God for adhesive!!.

I found a bloke in New York who makes spindles for PCB mill & drill. http://cnconabudget.com
You'll find detailed photos of the parts -- makes reverse engineering so easy.....

:-)

[Kevman]

hi all, ive been following all the diy spindle projects here with interest. a quick Q' do all of you run BESC's for your motors? ive been looking at hobbyking but their units dont seem to specify as besc or standard esc.

[njhussey]

As long as it's an esc for a brushless motor you're ok. If you're running it from the same power supply then a normal one is ok else if you want to run it from a separate 5V supply you'll need an "Opto" one.