[sploo]

I have a 3 axis CNC machine, running on servos. The machine is controlled using Mach 3, with a 45KHz pulse rate, and the control board (which runs from 5VDC) has 4 outputs (X,Y,Z,A), each providing ground, step, and direction.

The X, Y, and Z outputs are currently wired to servo motor drivers, which take a 24VDC supply, and are obviously connected to each servo motor (and take inputs for each servo's encoder).

I want to utilise the A output on the control board to add a 4th (rotary axis). This axis specifically needs to spin very slowly, and with plenty of torque. That seems like a job for a stepper rather than a servo; but I have no experience with steppers.

The few stepper drivers I've seen do take ground, step and direction inputs - so that appears physically compatible with my control board's output.

Ideally I'd need at least 7Nm of torque (approx 960oz in), so I guess a Nema 34 size would be required (e.g. 960 oz-in NEMA 34 Motor | Avid CNC | CNC Router Parts or https://www.cnc4you.co.uk/Stepper-Mo...15-4208-Nema34).

I could provide a new power supply just for the stepper, though it would be useful if its driver (and the stepper) would take 24VDC as it's what I have.

Does this sound realistic/feasible? Would there be any concern with the stepper drive with Mach 3 producing a 45KHz pulse frequency?

[ger21]

How fast do you need it to spin? A 960oz motor at 24V will have very little torque when spinning.

[CitizenOfDreams]

45kHz would be perfectly fine for a stepper driver. 24V would be a problem, as Gerry said.

So what is the advantage of a stepper in this scenario, other than the cost? Couldn't a servo motor "spin very slowly with plenty of torque"?

[sploo]

How fast do you need it to spin? A 960oz motor at 24V will have very little torque when spinning.

Thanks Gerry. Ideally no faster than 20rpm, and maybe down as low as ~3rpm.

I was originally looking at a 3000rpm 60W 24VDC servo with a 50:1 gearbox, followed by a 3:1 pulley (i.e. 150:1 => 20rpm max).

I have the pulleys and belts already, so could start with a 3:1 reduction on a stepper too (i.e. 3x motor torque).

I see the torque curve for the CNC Router Parts 960 oz in model shows that it actually delivers peak HP in the 400-1000rpm region; meaning it would likely be better to run it in that range and gear it down for max output torque (vs using it at slow speeds ungeared). That's assuming that's appropriate for a stepper (my experience is with servos).

What sort of voltage would be preferred for a larger stepper to get good torque; 48VDC, 80VDC?

[sploo]

45kHz would be perfectly fine for a stepper driver. 24V would be a problem, as Gerry said.

So what is the advantage of a stepper in this scenario, other than the cost? Couldn't a servo motor "spin very slowly with plenty of torque"?

Thanks. It's a torque and HP issue.

A 60W 3000rpm servo should produce around 0.2Nm of torque (at max speed). Assuming it's constant HP then if run at 1500rpm it'll produce 0.1Nm of torque, and only 0.02Nm at 300rpm (the torque should reduce linearly with a reduction in rpm).

If you instead gear the servo down by 10:1 for a 300rpm output (when running at 3000rpm) then it should produce around 10 times the output torque; i.e. 0.2Nm x 10 = 2Nm.

However, the datasheet for the CNC Router Parts 960 oz in stepper shows that it produces 3.6Nm of torque at 300rpm, so you're getting much more torque at lower rpm speeds without needing gearing; the downside is obviously that it won't spin at 3000rpm (should that be required).

For my application, I need a very low output rpm, so it seems sensible to go with a stepper + small reduction in gearing, vs a servo with larger reduction (given that low backlash gearboxes with high reduction ratios tend to be expensive).

[ger21]

However, the datasheet for the CNC Router Parts 960 oz in stepper shows that it produces 3.6Nm of torque at 300rpm, so you're getting much more torque at lower rpm speeds without needing gearing; the downside is obviously that it won't spin at 3000rpm (should that be required).

That's at 48V (Maybe 60V?) At 24V, you'd have roughly half the torque. I'd run it at 60V-72V.

[mactec54]

Thanks. It's a torque and HP issue.

A 60W 3000rpm servo should produce around 0.2Nm of torque (at max speed). Assuming it's constant HP then if run at 1500rpm it'll produce 0.1Nm of torque, and only 0.02Nm at 300rpm (the torque should reduce linearly with a reduction in rpm).

If you instead gear the servo down by 10:1 for a 300rpm output (when running at 3000rpm) then it should produce around 10 times the output torque; i.e. 0.2Nm x 10 = 2Nm.

However, the datasheet for the CNC Router Parts 960 oz in stepper shows that it produces 3.6Nm of torque at 300rpm, so you're getting much more torque at lower rpm speeds without needing gearing; the downside is obviously that it won't spin at 3000rpm (should that be required).

For my application, I need a very low output rpm, so it seems sensible to go with a stepper + small reduction in gearing, vs a servo with larger reduction (given that low backlash gearboxes with high reduction ratios tend to be expensive).

If you had the same size servo as the Stepper then the servo would be better, comparing a 60w servo with a 960 oz stepper the difference is of cause going to favor the stepper 60w servo is very wimpy and most likely not much good for anything CNC without some kind of reduction, even then you can't be very serious with your X Y Z cnc machine with 60w servos, they would work ok for a 3D printer

[sploo]

If you had the same size servo as the Stepper then the servo would be better, comparing a 60w servo with a 960 oz stepper the difference is of cause going to favor the stepper 60w servo is very wimpy and most likely not much good for anything CNC without some kind of reduction, even then you can't be very serious with your X Y Z cnc machine with 60w servos, they would work ok for a 3D printer

I apologise for the inadequacy of my CNC machine

It does indeed use 60W servos for the current three axes (with ballscrews). It only has a 39x25" cutting area, and happily manages 120ipm cuts at 1/4" DOC with a 1/4" bit in wood and MDF. It'll handle careful cuts in aluminium and brass, but it's not a large industrial bit of kit.

It's a fair point about the wattage comparison though. I guess as I've accepted I'd need a separate PSU for a stepper (for 48V or more) that I wouldn't be limited to the ~60W per axis region that my current PSU would impose.

Maybe I should just look at adding a fourth servo and gearing it down.