Ok, so I broke down and bought a high speed spindle from Wolfgang for my soon-to-arrive Taig. Since he just released a heavy-duty Taig mount for his spindles, the combined purchase was too hard to resist. :)

I'll be using the spindle to finish mill 6061-T6 aluminum using 1/32" cutters. The spindle includes a 540 can motor (brushed and inefficient) running flatout at 20K RPM, but I'd like to use a brushless motor instead.

The spindle is rated at 100K RPM, but I'd like to target around 40K RPM.

The brushless hobby motors used in RC airplanes look perfect, but I'm not sure they can handle the continuous operation (up to 12 hours continuous). I would suspect that the bearings would be the part that would fail over time, and I'd like to run it at 20K RPM using a 2:1 belt between it and the spindle, giving me 40K RPM at the cutter. Load would be light, since the 1/32" cutter would be eating at only about 15 IPM max.

However, I don't want to waste my money buying something that might be unsuitable. Has anyone else used a brushless motor like this with their CNC mill? Are any "warning bells" going off in your heads as you read this?

Thanks,
---Will

I havent used brushless motors on a machine yet, but i have been using them on model airplanes for about a year now and they run great.
It sounds like a good idea to put them on a cnc mill for hs spindle. the first brusless i bought was a year ago and i still use it about 2-3 times a week at the local park. after a 15-17min flght, the motors do become pretty warm. I'm only using 300 sized motors and they run at full power for most of the flight. Ive seen 600 size brushless motor used in large helicopters, these are the ones I would recomend. One concern would be the motor case, i know for a fact that some of the cases are made of carbon fiber, and if the motor is run for say 12 hours im sure the heat for defiinitly burn the epoxy in it. another concern is the controler, there motors have three leads, so you would need to build/buy a controller (maybe varialble speed). you wont be able to use an RC brushless controller they are ment to be hooked up to a reciever. hope this helps

Mike G

Look into proxon, they have a brushless dc dremel like tool that many have used with succes. I don't think you will get the desired torque out of a can motor and the bearings won't handle the load.

chris

Mike, the brushless motors I'm looking at have metal cases. That should help with the heat dissipation I hope. As for driving them, the controllers expect a .5ms to 1.5ms width signal to vary their speed. 10 years ago, I happened to have built servo driver using a 555 timer to generate the pulsewidths. It drove a servo just fine, and should drive the hobby-type brushless ESCs. What would be really cool is to get MACH3 to output a variable voltage that I could use to drive the ESC. A simple PIC (even a puny 16F84) could handle the voltage-to-PWM conversion for me. I'll look into it and report back when I have some results.

Look into proxon, they have a brushless dc dremel like tool that many have used with succes. I don't think you will get the desired torque out of a can motor and the bearings won't handle the load.
chris

Chris, thanks for the tip. I looked into the Proxon, first, and realized it was overkill for my needs. I'll be using the high-speed Wolfgang spindle to cut very small cavities in aluminum. No cavity is over 1/4" deep or over 1.5" in length. With an ultra-small stepover and stepdown of .00625", it should be hardly any stress on the cutter. I don't think a 1/32" cutter could handle much stress any way. I've heard those little babies snap off and go flying across the room at the slightest abuse. I'll definitely have an enclosure around my mill when it is whirring away at 40K RPM with a tiny cutter in the collet! :)

As for load on the bearings, the motor drives the spindle, so I'm not expecting any undo stress on the motor bearings. If the brushless motor shaft were the spindle, then I'd be in real trouble, wouldn't I!

Regarding torque, I wonder how I can calculate that? Does anyone know how I might determine the torque on a 1/32" 2-flute end mill running at 40K RPM, cutting .00625" into 6061-T6 aluminum at 10 IPM? I suspect it is negligible at that rate.

Mike, the brushless motors I'm looking at have metal cases. That should help with the heat dissipation I hope. As for driving them, the controllers expect a .5ms to 1.5ms width signal to vary their speed. 10 years ago, I happened to have built servo driver using a 555 timer to generate the pulsewidths. It drove a servo just fine, and should drive the hobby-type brushless ESCs. What would be really cool is to get MACH3 to output a variable voltage that I could use to drive the ESC. A simple PIC (even a puny 16F84) could handle the voltage-to-PWM conversion for me. I'll look into it and report back when I have some results.



Chris, thanks for the tip. I looked into the Proxon, first, and realized it was overkill for my needs. I'll be using the high-speed Wolfgang spindle to cut very small cavities in aluminum. No cavity is over 1/4" deep or over 1.5" in length. With an ultra-small stepover and stepdown of .00625", it should be hardly any stress on the cutter. I don't think a 1/32" cutter could handle much stress any way. I've heard those little babies snap off and go flying across the room at the slightest abuse. I'll definitely have an enclosure around my mill when it is whirring away at 40K RPM with a tiny cutter in the collet! :)

As for load on the bearings, the motor drives the spindle, so I'm not expecting any undo stress on the motor bearings. If the brushless motor shaft were the spindle, then I'd be in real trouble, wouldn't I!

Regarding torque, I wonder how I can calculate that? Does anyone know how I might determine the torque on a 1/32" 2-flute end mill running at 40K RPM, cutting .00625" into 6061-T6 aluminum at 10 IPM? I suspect it is negligible at that rate.

Thats not torque thats side load, torque is what the motor produces. Its most often referenced to horsepower but can also be used to refer to holding power of a servo/stepper. The rest is your feed rate, which really has nothing to do with torque. In which case your figuring chip load per flute and thats going to be very very small(at 12800 rpm give or take). I am not familiar with wolfgang spinldes, so my assumption is that it fairly small, and you have to attach the motor to the input of it directly(via a coupling of some sort). I know enough about about the hobby motors to tell you that they will not last much more then a half an hour at speed or they will burn up, even under forced air flow cooling(don't forget in an airplane its forcing air over the motor). I really think your best bet will be something along the lines of Proxxon or an air spindle. I dabbled in RC aircraft I know that the fellows that were teaching me did not like the electrics because they required a cooling down period.

chris

in2steam, I suspect things may have moved on significantly since you were involved with RC planes. People are flying 1/3 scales on electric now. I actually think there is mileage in running a brushless motor to drive the spindle. For example, I have a Hacker B50 motor rated to 100,000 RPM which will handle over 1kW with cooling. It will run indefinitely at lower powers, just be sure to run a well matched controller to the load so you don't overheat the ESC with switching losses through part-throttle running.
Chances are, you'll need something in the region of 50W of power to run those cutters based on my experience. I have a Proxxon IB/E spindle which is rated at 100W and slows down slightly when I push it with a 3mm cutter. Much more load than you intend.
I would question whether the Proxxon is not the way to go though. Even a relatively cheap motor/controller combo will cost you more than an IB/E or similar (they are about £70 in the UK).

If you belt it for a 2:1 ratio you are going to have side loads on the motor bearings.

Phil



The brushless hobby motors used in RC airplanes look perfect, but I'm not sure they can handle the continuous operation (up to 12 hours continuous). I would suspect that the bearings would be the part that would fail over time, and I'd like to run it at 20K RPM using a 2:1 belt between it and the spindle, giving me 40K RPM at the cutter. Load would be light, since the 1/32" cutter would be eating at only about 15 IPM max.

Thanks,
---Will

Here's a typical result from MePro.

1.0 watts. You could blow the material off with a mild cough.

endmill HSS
diameter (mm) 0.8
flutes 2
cut depth (mm) 0.4
cut width (mm) 0.8

surface speed meters per min 90
rpm 38000
mm per tooth 0.00409
mm per rev 0.00817
mm per minute 317
power required (Kw) 0.001
material removal rate (cc per min) 0.102

Phil



Regarding torque, I wonder how I can calculate that? Does anyone know how I might determine the torque on a 1/32" 2-flute end mill running at 40K RPM, cutting .00625" into 6061-T6 aluminum at 10 IPM? I suspect it is negligible at that rate.

Here's a typical result from MePro.

1.0 watts. You could blow the material off with a mild cough.

endmill HSS
diameter (mm) 0.8
flutes 2
cut depth (mm) 0.4
cut width (mm) 0.8

surface speed meters per min 90
rpm 38000
mm per tooth 0.00409
mm per rev 0.00817
mm per minute 317
power required (Kw) 0.001
material removal rate (cc per min) 0.102

Phil

Phil - Thanks! I see! 1W of power is consumed when cutting 1/64" deep at 12.5 IPM

Most of the hobby-type brushless motors seem to be able to produce 100W easily. This cheap 3800 KV motor & ESC looks about right, and the motor case doubles as a heat sink (no carbon fiber shell on this one):

http://cgi.ebay.com/3800KV-Brushless-Motor-w-30A-Brushless-ESC_W0QQitemZ330098578714QQcategoryZ2563QQrdZ1QQcmdZViewItem

Brief Specs:
Motor RPM/KV: 3800
Motor Max Watts: 250W
Motor Max Continous Current: 8-10A
ESC Minimum Voltage: 4.8V
ESC Max Voltage: 14.4V

So, driving it at 5V (and at 100% "throttle") will give me 19K RPM.

From the numbers above, at 5V the motor could generate a MAX of ~30W of power. So, each 1W consumed at 5V would require 200mA current.

I will only need 1W, so I suspect that that running virtually "unloaded" (the cutter will not be pulling many amps @ 12IPM, and 1/64" deep passes), and even with losses through the coupling system (misalignment, friction, heat, etc.) I would be drawing WELL under 1A @ 5V, with 250mA probably being typical for my application.

How's my math, guys? Did I goof up any where?

---Will

Be careful. The program doesn't say but I think the calculated power requirement is at the tool, not the input. How would the program know your spindle drive efficiency.

Regards
Phil

Phil - Thanks! I see! 1W of power is consumed when cutting 1/64" deep at 12.5 IPM

Most of the hobby-type brushless motors seem to be able to produce 100W easily. This cheap 3800 KV motor & ESC looks about right, and the motor case doubles as a heat sink (no carbon fiber shell on this one):

http://cgi.ebay.com/3800KV-Brushless-Motor-w-30A-Brushless-ESC_W0QQitemZ330098578714QQcategoryZ2563QQrdZ1QQcmdZViewItem

Brief Specs:
Motor RPM/KV: 3800
Motor Max Watts: 250W
Motor Max Continous Current: 8-10A
ESC Minimum Voltage: 4.8V
ESC Max Voltage: 14.4V

So, driving it at 5V (and at 100% "throttle") will give me 19K RPM.

From the numbers above, at 5V the motor could generate a MAX of ~30W of power. So, each 1W consumed at 5V would require 200mA current.

I will only need 1W, so I suspect that that running virtually "unloaded" (the cutter will not be pulling many amps @ 12IPM, and 1/64" deep passes), and even with losses through the coupling system (misalignment, friction, heat, etc.) I would be drawing WELL under 1A @ 5V, with 250mA probably being typical for my application.

How's my math, guys? Did I goof up any where?

---Will

Will did you see this thead. Probably worth a look.

http://www.cnczone.com/forums/showthread.php?t=23097

An Aussie builder put one of those same spindles (I think) on a machine with the same sort of motor as you are looking at.

Greg

The little brushless motors for RC cars do lose bearings pretty quick. They are rebuildable and all, but a couple of hundred watts out of a couple ounce motor and something has to give. They are designed for maximum power out of a very small and light package, not longevity. Many of them require rebuilding every 10 hours of run time or less.
If you want a brushless DC motor, with speed controller built in no less, that will handle high side loadings and long runs, try something a bit larger. Those annoying little electric scooters, and electric bicycles as well, have 250W-400W brushless DC motors that have bearings that can handle a LOT of abuse. A PWM speed control board is integrated into the motor housing on many models. Nice bonus is that they often have a brake circuit on the motor as well. (useful for e-stop!)
You can get them brand new uninstalled >very< cheap off of EBAY as the fad is over for scooters. $20-40 bucks will get you a very nice motor.
They are light, but are still 4"-5" in diameter instead of 1", but no worries about overheating or bearing life....

in2steam, I suspect things may have moved on significantly since you were involved with RC planes. People are flying 1/3 scales on electric now. I actually think there is mileage in running a brushless motor to drive the spindle. For example, I have a Hacker B50 motor rated to 100,000 RPM which will handle over 1kW with cooling. It will run indefinitely at lower powers, just be sure to run a well matched controller to the load so you don't overheat the ESC with switching losses through part-throttle running.
Chances are, you'll need something in the region of 50W of power to run those cutters based on my experience. I have a Proxxon IB/E spindle which is rated at 100W and slows down slightly when I push it with a 3mm cutter. Much more load than you intend.
I would question whether the Proxxon is not the way to go though. Even a relatively cheap motor/controller combo will cost you more than an IB/E or similar (they are about £70 in the UK).

Actually June of last year, I had a smaller park flyer untill it was reduced to small pieces of foam and plastic. The prospect of replacing $100 of plane every once in awhile was not at cool as it was as first. I decided to keep with my live steam hobby instead.

A dynamic load(increases relative to speed) such as a prop is entirely different then a static(just increases) load like a cutter. I agree that you will need next to nothing for load, either way the chips will be well, dust. The other thing you are not thinking about is the heat from the cutter, at those speeds even a light load will require air and or water cooling. It would be interesting to see, as the smaller motors are compact enough, I still think the duty cycle will not allow as long a run as you think. As for using a belt at those speeds to reduce 2:1 I can think of very few price reasonable materials that will do for the long term.



As for the electric scooter those don't spin anywhere you need them to....

chris

As for the electric scooter those don't spin anywhere you need them to....

chris

For direct drive that would be very true, but with the stock pulley system they will do quite well. The scooter motor has the same RPM as my original Taig AC motor did, only lighter and more powerful and with a broader speed range, and fit in the same mount using the same original Taig drive pulley. They are very similar to the brushless treadmill motors others have used, just usually in lighter aluminum cases and with speed controls built in.
I have an old style belt-driven high-speed dental pencil handpiece mounted to the t-slots on the front of the spindle block, and the stock belt moved over onto it runs the thing much faster due to the small pulley on the handpiece (around 12,000 RPM is where I usually use it) allowing for PCB and engraving as well. I just switch back and forth between that and the stock ER16 based on need, simply by moving the belt a couple of inches to either driven pulley.

For direct drive that would be very true, but with the stock pulley system they will do quite well. The scooter motor has the same RPM as my original Taig AC motor did, only lighter and more powerful and with a broader speed range, and fit in the same mount using the same original Taig drive pulley. They are very similar to the brushless treadmill motors others have used, just usually in lighter aluminum cases and with speed controls built in.
I have an old style belt-driven high-speed dental pencil handpiece mounted to the t-slots on the front of the spindle block, and the stock belt moved over onto it runs the thing much faster due to the small pulley on the handpiece (around 12,000 RPM is where I usually use it) allowing for PCB and engraving as well. I just switch back and forth between that and the stock ER16 based on need, simply by moving the belt a couple of inches to either driven pulley.

I don't knwo what kind of motor you are using but I have two for another project and they don't have the same size shaft, its not keyed and they only about 1/10 HP opposed to 1/4 hp. I suppose if you got a self contained unit it would work, the only other problem is that the original controlers used a hall effect sensor for speed control so if its intergrated that could be a problem. The two I have are plain motors, nothing real special.



chris