Cnc Spindle Project

Hi i’m planning a iso20 spindle(12k rpm max) and i would really like to avoid the cost of an industrial servo System (only ~1.5nm @10k rpm for ~400€+ . That is still more torque than my 2.2kw china spindle? I think).

That is why i’m trying to find out if a odrive/rc bldc + pulley is the right combo for me . The Problem is i’m clueless when it comes to rc motor stuff :neutral_face: i hope that maybe someone can help me .


  1. Min 4nm or more ? (as long as the motor got enough rpm i dont mind using a pulley)

  2. Works with my 48v, 30a, 1600kw powersupply (got ~200w headroom)

  3. Max 12k rpm on the Spindle

Greetings from Germany

Ok, so for a start, no motor/drive combo is going to produce 4Nm @ 12k rpm at the same time with a 1.6kW supply, because 12k rpm = 1256 rad/s, and if you multiply that by 4Nm you get 5kW mechanical power. Obviously, conservation of energy applies and you only have a 1.6kW power supply.

That said, if you don’t need these specs simultaneously, you could pick a motor such as this:
Given that kV = 172 rpm/V ~= 0.058 Nm/A (by my dodgy calculations), at 80A you would get 4.65 Nm. ODrive can just about handle 80A continuous without exotic cooling, so that should give you the low-speed torque.

For high speed you need to be aware of ODrive’s limit of 600Hz (iirc) electrical. This motor has 12 pole pairs, at 600Hz electrical that equates to 18k rpm max no-load speed (commutation limited) (i think)

However, at Kv=172 and 48V you only have 8256rpm no-load speed (voltage limited). So despite the ODrive’s limitation, the back-EMF of this motor at 12k rpm is more than your supply voltage, so you cannot achieve that speed.
Also, there is a second limitation of ODrive that modulation magnitude is capped at 80% because of the SVPWM scheme (iirc) so this voltage-limited speed drops to only 6k rpm.

If you picked a similar 5kW 12pp motor wound at 350 Kv, you could get 12k rpm, but only 2.3 Nm at 80A. So you could up that to 120A (the very maximum of ODrive) and maybe you could do 3.5Nm with some fancy (liquid) cooling, but you are pushing the limits of permanent-magnet motor physics here, never mind ODrive :slight_smile:

1 Like

AC Induction motors are nice when you don’t need the speed and torque at the same time, and ODrive supports ACIMs. Though it may be quite hard to find 48V ACIMs in the 2-4kW range.

it might be possible to find a suitably heavy 3-phase transformer or variac that would allow the ODrive to run more common 415V ACIMs? :grin:

hahaha at that point I would just open the motor and start rewiring terminations from series to parallel.

1 Like

Ive come up with a couple of these designs. Your biggest problem at high power is going to be heat. As these motors were designed for Short operation at high airflows(ie. Planes) you`re going to overheat the motor pretty easily. Your best bet is to water cool the motor and use belt reduction to gear down the motor. This icreases the torque at the cost or rpm, which is ususlly what you want to get a usable surface footage for the machine youre using.

6k rpm of usable torque on a homemade machine is a lot better than 12k of low torque that the machine can’t feed fast enough.

1 Like

Hi thx for all the replys so i guess it’s going to be the odrive with a Server psu ( will prob blow my 16a fuse :expressionless: ??) or a servo with 3nm @100rpm and ~1.06nm @12k rpm. Btw i meant 4nm at low speed around 1-2nm at 12k is still way more than any china spindle can do i think ( atleast the ones under 4kw) .

And i guess it’s safe to say that my 1.6kw 30a psu is useless in my case right ?

Anyway is there a “safe” way to use a Server psu or should i just go with the servo/drive combo(~1.8kw ) that does 10k rpm right out the Box without a belt drive ( and runs at same price atm with the high chip prices)?