View Full Version : DIY ATC Spindle

22-04-2019, 05:54 AM

I am building my own atc bt30 spindle. For the motor i will use my current 2.2kw water cooled spindle with 1:2 gear ratio for more torque at lower speeds but since it is heavy, noisy and take a lot of space in my z axis configuration (no space for air cylinder) i was thinking to use smaller brushless motor for rc boats. Since i dont have that much electrical knowledge i would like to know if this motor would even work for such task? They are small, high rpm, water coolled but i have no idea if this can handle 8h+ work a day? Here is the link for the motor:


Please help?

22-04-2019, 07:11 AM
Without any real knowledge on my part (so typical for an internet response) - you might want to ask the supplier for the run-out figures (both axial and radial) and guaranteed life of the bearings used. Those would be my main concern, and it's these mechanical (rather than electrical) characteristics which would influence the use of such a motor in this application.

Regarding the electrical side of things, you have to factor in a low-voltage DC supply at whatever power-level you anticipate, control system (stick with the RC controllers - they are designed for this), and heavy duty power distribution (cabling, connectors) as well as cooling systems (rough rule of thumb : for a given power, the higher current/lower voltage systems will have higher power losses in the form of heat).

I can't help but think that you're trying to re-invent the wheel here where solutions already exist, but you have my grudging respect to try this. I'd be interested in hearing how successful you are - particularly if you get the ATC working.

22-04-2019, 10:16 PM
Here is the section view of the bt30 spindle setup. All the parts for spindle are made or ordered. Unfortunatelly i didnt think about diy spindle when i was designing the cnc machine, so now i am facing lack of space on z axis.


Sole purpose of the motor would be to run the spindle shaft.

23-04-2019, 09:04 AM
btw, if anyone is interested, i could make another shaft. (shaft, pull bar and lock socket)

23-04-2019, 06:00 PM
You need to look at proper spindles and use a stack of conical spring washers to retain the tool via pull studs and a readily available, off the shelf puller (petal/clamp/claw are useful search terms)
You don't have space in that design for a good pull system.

23-04-2019, 07:54 PM
I dissagre. This system works perfectly. There is a stack of 42 belleville washers in the shaft, with enough force to hold the tool with six 5mm bearing balls. I will demonstrate when the spindle is build in a month or so.

23-04-2019, 07:58 PM
Yes, it will have the tension, initially.
But there is a reason the spring stack is much larger in commercial spindles, they don't rely on what works, they rely on what doesn't fail prematurely in service.

23-04-2019, 08:07 PM
that might be true, but even if the springs fail, it is much much cheaper to repair the damage than buying 3k+ spindle. At the end of the day, this is still home made cnc, not proffesional cnc so i hope it will do the job.

23-04-2019, 08:13 PM
10krpm spindles are closer to 1k.
All you'd need to do is use a larger tube & bearings to accommodate adequate components, why would you not?

23-04-2019, 09:01 PM
i want 15krpm+ spindle with atc and must fit my z axis configuration. Got made mine for almost half the price.

05-09-2019, 12:12 AM
I like your approach. I am willing to help with calculations and design. All you have to do is point me at the desired components and define your space constraints

you can reach me at [email protected]

John McNamara
05-09-2019, 08:23 AM
Hi Husky
Hi Irving (I sent you an email Irving)

Gee Husky sounds like you are on the right track to me. Belleville washers do fail after millions of operations, a common service repair for spindles in commercial use. For home use they should last a lifetime. And making something is always more satisfying than off the peg.


09-09-2019, 12:41 AM
Im looking to do something similar. I'd be interested to see more details on the spindle size and bearing arrangement.

Is it a squeeze release or does it load the bearings when pressurised.?

Is the spindle hardened and ground?

26-12-2019, 09:41 PM
Sorry for late reply, been busy with work. My diy BT30 spindle works great, couldnt be happier. 10h/day for few months now, no serious issues yet. Cut like a butter but the lack of power (2.2kw spindle) is holding me back to cut like a pro. So i am already considering upgrading motor with 3kw spindle or 3x more expensive servo motor with 11Nm ofcl torque.

Ross77, it is sqeeze release, i have a steel bar to sqeeze the washers. No intention to make it fully automatic, no need for this yet, and a lack of room to place cylinder over the shaft. The spindle shaft is hardened and ground. Here is old vid of the spindle. Runout is approx. 0.002mm, tool holder and collets are from china, thats why such runout in vid, which is still great for my first diy spindle


29-12-2019, 01:34 PM
That looks about 30um / 0.03mm runout and presumably worse at the cutter tip. That means the cutter will be doing most of the cutting on only one or two edges. You should really use a better quality toolholder, or you will be unable to do any serious cutting without breaking tools and / or poor surface finish.

Rather than use an ER collet toolholder, you should get some (Chinese) sidelock holders that are inherently much better due to the lack of collets - it's the collets that are the biggest contributor to runout. Then you can see how good your spindle is - and make best use of it without breaking cutters.

Very interesting work - please let us know how you are getting on!

31-12-2019, 06:14 PM
Ross77, it is sqeeze release, i have a steel bar to sqeeze the washers.

Surely that makes it an MTC? ;-)

01-01-2020, 11:36 PM
I will test better quality toolholders later and see how it goes. In general, i am very happy with the build, next upgrade is 3k spindle for a motor, or more expensive servo motor like 3.5kw fagor with 11Nm of torque. With 1:3 ratio for higher rpm it can cut serious chips.

I wont upgrade to fully automated tool change anytime soon although its a simple upgrade with diy cylinder over the shaft. I would need to modify the cnc frame as well, so the toolholders would be away from the flying chips, it need extra care to prevent chips getting between the toolholder and shaft cone. The table is usually covered with thick layer of chips, its a mess. Much faster to change tool manually.

19-02-2020, 09:41 AM
I am getting BST Automation 3.8KW, 15Nm servo motor for my spindle shaft motor, with 1:4 ratio to get 10krpm. Hope this will work much better than current 2.2Kw water cooled spindle providing only half the power of the new setup and with much much less noisy working.

19-02-2020, 11:04 AM
Butterfly air ratchet with standard modification of valving moved to an easily accessible area would have removed the need for springing, or a lever, for a Non-ATC spindle.

19-02-2020, 10:19 PM
Yes you can use a threaded drawbar with an ATC system. The “drive dogs” shouldn’t actually be doing any torque transfer as such, so they can safely be removed. That means you don’t need to index the spindle for tool changes. That’s how the Fadal ATC works on my Shizuoka AN-S. Makes the system design so much easier...