Motor advice for a machine tool

Joined
Feb 16, 2016
Messages
6
Hi,

I'm making a mill-turn center that needs a lot of motors to work. I know it's slightly off topic from electric vehicles, but the mill turn center will be used to manufacture custom electric vehicles, so...

Requirements;
-The main spindles of the machine need to be able to run in a speed range from 0 to around 3,000 RPM.
-The spindle will need 2kW or more at higher speeds when cutting metal.
-Low end torque preferable.

My plan is to use a low KV outrunner motor with a built in hall effect sensor, and drive it with a high voltage electronic speed controller powered by Lipos. For initial testing, I'll control it with a cheap servo tester. Eventually I'll control it with a CNC control board PWM signal. I have special absolute rotary encoders and I'd like to be able to do closed-loop position control using the outrunner motor eventually.

My main concern now is that I'm not that familiar with BLDC motors and controllers like most of the people on these forums. I'm fairly certain I need a hall effect sensor and a low KV BLDC motor, but the details and pitfalls are unknown to me. I'd like to be an expert on this subject eventually, but for now I'm new to it. I have access to 3d printers and CNC machines to make what I need; I can build some of my own hardware.

So, can anyone give me advice on what to buy for my project? I'm leaning toward this motor to start with;
http://alienpowersystem.com/shop/brushless-motors/6374-sensored-outrunner-brushless-motor-130kv-3000w/
 
Interesting. I don't think that's the right kind of motor for the job. You may have noticed that machine tools have physically large massive motors for their power ratings. That's beneficial for torque, heat management, rigidity, vibration, and bearing longevity. If you use hobby vehicle motors, you have to contrive your own solutions to all these issues.
 
I've been using a re-wound 80mm outrunner with a home made main shaft that will chuck a 3/8" end mill
I do have a angular thrust bearing in the mounting plate to reduce the stress the motor ball bearings see in "plunge"
FN CNC.jpg


it does ok....I am in the middle of a rebuild adding better supported main rails on the x axis.
machines soft materials great....but I had to baby it cutting aluminum, not from lack of motor power. just too much flex in the table supports with unsupported 30mm shafts.
 
Thanks for the quick replies! The main turning spindles can either be belt driven or gear driven, either way there will probably be a speed reduction of about 50% to double the torque. There are a bunch of reasons I want to use a hobby RC motor to drive this machine, the obvious ones being cost, weight reduction, and being able to use a soda-can sized motor; it's a sliding headstock machine, so accelerating 1 or 2 kilograms of BLDC motor is much preferable to accelerating a 30kg AC motor.

The machine spindles have two 7209 45mm ID angular contact bearings supporting a steel tube as the spindle, and the belt pulley will have it's own separate bearings, so the electric motor will never see radial or axial loads. It will be fan cooled at all times to prevent heat buildup. I have a clutch mechanism to engage or disengage a large stepper motor to the same spindle, this will be used for rigid indexing/rotary positioning and tapping. Eventually I'd like to remove the stepper and have the BLDC motor do all of the work.

Can anyone recommend ESCs, batteries, or anything else I would need to get this machine spinning? I don't want to buy the wrong kind of motor or controller.
 
I'd consider using this: http://vedder.se/2015/01/vesc-open-source-esc/

See: https://endless-sphere.com/forums/viewtopic.php?f=31&t=73234 et al.

What is your torque requirement?
 
As far as torque goes, more is better. I'm using the assumption that 1 horsepower for 1 minute can remove 1 cubic inch of steel, or 3 cubic inches of aluminum.

A 2kW cut at 2,000 RPM would mean about 9 N.m of torque, which is in the upper range of what I'll be doing.
 
I put the specification (as posted on the Alien website) of the motor that you linked to, into the motor comparison spreadsheet. It's somewhat underwhelming, to put it mildly...
 
Woah, just looked at the spreadsheet. 1400 watts of power output but 1000 watts of heat? I guess it was too good to be true.

Is it some kind of design flaw in that specific motor, or is it just inherent that low KV motors will be less efficient? I may have to go with a larger speed reduction in the pulley/gear if that's the case. At 1kW of waste heat, I might as well just duct tape a toaster to my machine spindle.
 
Well, it could be a typo in the resistance figure given but probably not a simple decimal point displacement. If not, look elsewhere would be my advice.

The velocity constant has absolutely nothing to do with efficiency.....
 
Miles said:
Well, it could be a typo in the resistance figure given but probably not a simple decimal point displacement.
Well, I suppose it could be 56 milliohms........

http://www.hobbyking.co.uk/hobbyking/store/__18184__Turnigy_Aerodrive_SK3_6374_149kv_Brushless_Outrunner_Motor.html
 
Back here again a year and a half later for some updates.
overview1 july 16.jpg
turret1 july 16.jpg
I ended up using a Turnigy 6374 motor with 149kv for the main turning spindle, and a 45mm 510KV motor for the milling spindle. Neither one is sensored. I used two 12v server power supplies in series to give them up to 24 volts and up to 60 amps.

The electric stuff worked decently, but the mechanical connections were bad. I tried to transmit torque with a set screw since the motor shafts are so small. I should have gone with the APS motor since it has a 10mm keyed shaft.



Now I'm building a second version of the machine, and it must be a lot more powerful and reliable. So I'm going to buy the 63mm alien motors for both the turning and the milling spindles. I just wanted to run a few numbers by some of the experts here.

First of all.... How much torque can a keyed 10mm shaft transmit in foot-pounds? I don't know how hard the steel is on RC motor shafts, I'm guessing around 50 Rockwell C.

The 6374 alien motors are advertised as having a bit over 3 kilowatts. They don't specify if that's peak power, at what RPM, or how long they can maintain that, or if it's input or output. I'm skeptical that they can hit that power at all. Someone enlighten me please!

But assuming they can do that just a second, it means we can approximate their torque. A 6374 90kv sensored outrunner that can do a full 3200 watts would reach it's max power at about half of it's max RPM. So 90kv * 48 volts = 4320 RPM maximum, half of that (rounded down) is about 2100 RPM. Doing a quick power-torque-rpm conversion, I'm getting a torque output at that RPM of just over 15 newton-meters, or 11 foot-pounds. Since the max power is at half the max RPM, and the torque decreases linearly with rpm, that means the maximum possible torque this motor can output at 0 RPM would be approximately 22 foot-pounds.

Is my estimate process correct? Can someone please give me guidelines for how much power and torque these motors are capable of, and how long that can be maintained? I know some of you on this forum have these very motors on your longboards.

My new machine will have chevron-toothed timing pulleys at a 1:1 ratio. I'll probably be using the VESC as the main motor controllers, possibly the new ODrive to test outrunners as servos. Encoders on motor shafts as well. a stack of 4 hot-swap server PSUs to supply 48 volts at 60 amps or so; the power supply can be tapped for voltage at 12v increments.
 
https://www.teknic.com/products/clearpath-brushless-dc-servo-motors/
I'd strongly suggest looking at these. You'll spend a bit more than a couple of turnigy motors and controllers, but these have everything built in, and would let you get rid of your stepper at the same time, and their built for this exact kind of application. I'll be building myself a fixed gantry cnc sometime and these will be my first port of call for motors, including for a 4th and 5th axis.

your calculations for torque are probably well over what they're actually capable of - the 80100 size are known for maxes around 10-15Nm, or ~10ftlb. a 150kv motor will produce ~0.065Nm per (phase) amp, so to get the torque numbers your talking you'd need in the region of 450 phase amps - I probably dont need to tell you that the motor will cook almost instantly at that current level (and it will saturate long before then too, so you still wouldn't get the 20ftlb your after :( )

the best way to estimate torque output is:
9.8/kv (9.8 is an approximate conversion value for the motor constants to go from kv to kt, or rpm/v to Nm/A)
so 9.8/149kv=0.065

desired Nm/answer to above = amps needed for your desired torque (ignoring saturation)
so 29Nm/0.065=450A

generated heat is then (amps^2)*(Phase Resistance)
so 450^2 x (i dont know the phase r, but probably around 20mOhm at a wild guess) = 4000W of heat.

this heat generation value gives you some indication about if your desired torque levels are reasonable - more than a few 100w in these little motors means it wont do it for long.

Sorry didn't want to rain on your parade too much mate, but i really think it'd be easier (adn possibly cheaper) to go with some purpose built motors from teh get go... That said i like the look of your machine - any vids of it in operation yet?
 
My newer prototype has axis motor mounts with multiple bolt patterns to be able to mount NEMA 23, 24, and 34 motors; all you have to do is switch the coupler on the motor shaft. I've been planning for a while to use both steppers and the Teknic servos, my main problem is that this machine has so many axes that using appropriately sized Teknic servos on every axis would cost more than the rest of the machine. So I'm leaving it open to use steppers, Teknic servos, and/or O-Drive servos once they're a bit more developed. There is a company called CUI that makes various small shaft encoders that would be perfect for this, they mount easily.

Sadly, nobody makes decent affordable spindle motors for this type of stuff, which is why I'm using RC motors. The spindles will have their own P4 angular contact bearings at 45 and 65mm ID's, so the motor shaft will only have to take the force of the pulley, I'm trying to put a reinforcement bearing in as well to take some of the pulley load off of the motor shafts.


Based on what you said, I will need to use lower KV motors and limit high current to a few peaks. The power supply is a stack of four server PSUs, so I can do (theoretically) 3000 watts at just over 60 amps. Regular household power outlets are 120v 15 amp, which limits them to about 1500 watts for practical use before they trip breakers. But the 240v appliance outlets work with the PSUs as well, which will be adequate to drive all four PSUs at max load.

Using the 9.8/KV formula with a 6374 90KV motor, a maximum of say 50 amps to the spindle motor VESC would do (9.8/90) * 50 = 5.44 Nm = 4 ft-lbs. I hope that's within the saturation current.

The 90KV motor would limit this spindle to around 4000 RPM, but I'll only be using half of that with a spinning lathe chuck. The milling motor would need a higher KV, perhaps 130.

The peak of 4 foot pounds is kind of disappointing; it means I would only be able to take skim cuts on a 3 inch radius part. I guess I can use bigger motors and have a 2 speed pulley to compensate.

I have two videos of the mill-turn first prototype, here's the second one;
https://www.youtube.com/watch?v=1GkhRZA5BUY
 
Back
Top