3D machining, servo or stepper

Hi. Having recently bought a CMM (Coordinate Measuring Machine) with the intention of converting it for 3D machining, I am seeking advice on whether to keep the servomotors, or replace them with steppers. The existing system has motors with tachos, a toothed belt (approx) 8:1 speed reduction, and a strip along each axis for position sensing. Y and Z axis travel is about 900mm, X axis to be determined as the original granite table went elsewhere, but will be about 1.5m.
I am thinking of using a Geeetech GE2560 controller with high-power motor drivers, but the supplier seems quite reticent about supplying a schematic for the board. L298 drivers would probably follow.

Thanks in advance

Please post a link to the datasheet for your servomotors. I suspect you are not referring to the types of hobby servo usually used with Arduinos - the sort used to control flying model airplanes.

The suitability of a motor for a CNC task depends on its precision and its torque as well as the rigidity of the frame it is attached to.

What sort of material are you planning to machine and what size of material are you planning to work with.


The original system sat on air bearings on a 1.5 x 0.900 granite surface and weights about 300kg

A mill needs to be physically strong and stiff, a position measuring system much less so as the contact force is as close to zero as makes no difference, whereas milling has huge forces. Or are you only intending to machine foam?

A CMS uses closed loop systems for accuracy.

Most CNC machines also have closed loop systems but for home use it is generally not needed on the CNC machines

I would suggest build one CMS machine for one task and seperate CNC for any machining work.
As mentioned the stresses are completely different and so are the tasks.

@sbkenn, an EBay link is not a datasheet. The datasheet is produced by the manufacturer and describes the motor's technical capabilities - torque, speed, voltage, current, forces for which the bearings are designed, etc etc


L298 drivers would probably follow.

That, in my opinion, would be the very last choice for a stepper motor driver. They are DC motor drivers and crappy ones at that. Their ancient technology is very inefficient. Modern stepper drivers are much much better and offer current control, micro-stepping, high efficiency and more.

I just realised (re-read) the OP's quest.
Hate to even think of the cost of heating a granite bed for use as a 3d printer bed !
Or the time needed for that task.

Where did he say 3D printing? I see 3D machining.

This is the machine. I won't be using the granite surface, and the base-plate will move rather than the (300kg) gantry. Ferranti Coordinate Probe Machine & Granite Bed. As for the L298, that is the driver/power amp rather than the controller, and the efficiency and microstepping is the business of the controller.
I did say 3D machining, conventional and additive.

This is the machine.

All the pictures are obscured by the "bidding has ended banner". Also the size is not at all clear. Can you take some photos yourself and post them? Include something in your photos that will give us a sense of size.

You never answered my question about what material you want to machine or the size of the piece of material.

If the machine is designed for precision measuring I would be very surprised if it could withstand the forces needed for CNC cutting.


The machine had a working volume of about 1400 x 900 x900mm and consisted of a gantry which floated on the granite, a cross slide floating on a 250high x 200wide beam with 4 ground corner faces , and a vertical bar of about 75mm square. The system is dismantled at present. The only significant modification will be twin rather than single belt control of the main table

A CMM is just that.
It is not designed for anything else.
If it is dismantled it will of course need to be rebuilt and re-certified / re-aligned to be used as a CMM.
Its use as anything else will void its status / calibrations as a CMM.

More so as it was a commercial / industrial unit !

You really need to take these two seperate tasks as two different projects.

I have no intention of using it as a CMM again, and IMO, it is as rigid as many smaller non-production milling machines.

Let's pretend that the machine is rigid enough to make a milling machine capable of light work.

Now let's go back to the Original Question - the existing servo-motors vs new stepper motors.

I'm assuming, whichever motors you use that you want to control the system with an Arduino.

Do you have a control interface for the servomotors? I already asked you to post a link to the servomotor datasheet, but you have not done so. Without a suitable controller, and the datasheet that describes how to use it, it is hard to see how you could use those motors with an Arduino.

The other reason I asked for the datasheet for the motors is that the work required from a motor driving a milling machine is likely to be a lot higher than for a measuring machine. You need to know how much torque and speed will be required for milling and then consider what motor would be suitable.


OK given that we have cleared up the confusion there are new considerations.

First being the size of the machine.

When moving into that size of machine you would want larger stepper motors as the existing is NOT designed for the loads you will be introducing.
That sort of size would required at the very least some NEMA 24 but preferably 34 series.
They will pull a lot more current than the shield you propose.

That takes you into the TB6xxx series at the very least which rules out the board you selected and moves you into the MACH series controllers.

It may also rule out the belt driven approach too or at the very least force a reconsideration in the type and size of belt to be used.

The steppers and drivers can often be had at a reasonable price as a kit leaving it up to you to decide on the drive system.

With that size of machine I would certainly consider a ballscrew approach rather than belt or acme screws and probably at the very least 16mm but preferably somewhere in the 20mm range.

You also need to take into consideration power supply requirements but you may be able to pull something from the existing system for that aspect.

Given the size of the machines I would probably have taken something "off the shelf"

Stepper kit

I dont know what you paid for the CMM but it might have been a poor choice as a starter project.
My next machine is going to be a 6090 or slightly larger.

Thanks for the input and sorry about the delay in responding, I have had some health issues. I have also just ]partially reassembled the machine.
I appreciate that the milling loads would be higher, certainly higher than most DIY machines. The current DC motors are 0.2 N.m/28 Oz.in, times pully gearing of 20 : 150 gives me an output torque of 1.5/210, then x 5 for torque -> force @1.5cm radius -> 100N pull, if my calculations are correct. All belts are steel reinforced PU toothed. I don't see how to attach images, so

The cross beam is 50", weight approx 150kg, the R.H. "leg" about 80kg

I paid £150 for the whole thing which included a Renishaw "wrist" and probe which are trading for up to £2k on eBay. I have no use for that part.

Power supply is half a tonne of batteries connected for 24V ! I live on a small ship.

I don't see how to attach images,

Simple Image Posting Guide


I have no intention of using it as a CMM again, and IMO, it is as rigid as many smaller non-production milling machines.

That means it weighs around 2 tons or so?

Big milling machines are meant to handle large forces (10's to 100's of kgf) without distorting more than a few microns, this stiffness is usually achieved by using large amounts of metal (cast iron usually) in appropriate geometries.

You'll probably be able to mill very soft metals and wood with this thing but I suspect it will kick back if you are more ambitious. Under-stiff mills are dangerous as they can grab workpieces and propell them across the room like shapnel when the tool digs in.

I still have the software for the Renishaw here.
It must be about 15 years old by now.

Used it to test the runout on machines after any changes to the leadscrews , spindles, etc,
It was a great little portable device to use with a laptop.

IIRC the BLUM laser system suceeeded it although I only ever installed one of those for tool changeover measurement.