[svenakela]

I am sourcing parts for a large machine and because of a tight budget we need to go for large steppers instead of our usual servo's.
I have no experience in steppers. When looking at different sources like web shops, eBay, catalogs etc there is a pretty big difference in pricing.

In general terms, are there much differences in different steppers?
I mean, steppers are steppers. Or?..

[steve123]

The bigger steppers have higher inductance. This greatly reduces torque at higher RPMs.
Here is a calculator: File:Motorcalc.xls - TAMI
These guys have a pretty good 960 oz/inch nema 34 motor: Nema 34 CNC Electronics | CNCRouterParts
This motor has 2 ohms inductance. The corner speed (where the motor starts loosing torque) is only 614 RPM, but that is doing well for a larger motor!
For comparison, The 906 oz-in KL34H295-43-8B in motor here Page Title has an inductance of 3.3 ohms. It's corner speed is only 372 RPM!
I would go with the Gecko 201x drivers. Cheap Chinese drivers will not give you the same performance.
This kit is worth considering: 4-Axis DIY Nema 34 Kit | CNCRouterParts
-Steve

[svenakela]

Thanks Steve!

I did my own calculator sheet after the Gecko formula, and one funny thing is that a higher pitch on the screw gives a smaller motor.
It's here if you want to try it: http://bit.ly/16DrFdP (The odd numbers in the yellow fields are only to make the same calculation as the Gecko example).
I was aiming for Leadshine drivers. They are used by Tormach (I've been referencing them quite a lot for this build) and after their testing they ended up with a very well written report.

[MooseCat]

I mean, steppers are steppers. Or?..

I've looked through many data sheets in recent times. My impression is that the bottom dollar chinese motors (eg ebay, web shops) often use invented numbers for values like torque. If you are lucky you will get ~90% of the stated torque, but 60-80% is more likely.

[svenakela]

I've looked through many data sheets in recent times. My impression is that the bottom dollar chinese motors (eg ebay, web shops) often use invented numbers for values like torque. If you are lucky you will get ~90% of the stated torque, but 60-80% is more likely.

That's the difficult part. Same goes with servos. Figures are usually on the sugar side.
I have a machine with surprisingly small Nidec servos, No one thought they were capable to run that pretty heavy and large machine. I had to cut down the acceleration to 2500mm/s² because the base was shaking too much. They are high qually motors that really have the figures they are rated with. I am pretty sure no names need to be bigger. of course that must go with steppers as well.

[ger21]

Can you be more specific about the application.
All steppers are not created equal, but it's not that some are better than others. They are made with different current and inductance ratings, which will vary the torque curve depending on the drives and voltage you're feeding them.

In most cases, the higher the current and lower the inductance, the faster you'll be able to spin the motor. The downside is that you may need to spend more money on drives to achieve this higher speed. Higher inductance motors can usually use cheaper drives, if the rpm required doesn't exceed the lower capabilities of these motors.

[svenakela]

Can you be more specific about the application.

Sure I can but it's not what I want to discuss, more like the rest of your post which was very informative.
It's a small but heavy machining center, will be 5 axis in the end with a tool changer. It is going to be built with a small floor foot print with a working area size of XYZ 500/500/400 mm.

We've had a discussion about the same thing on the LinuxCNC mailing list and the rough Gecko formulas I used are just that. Rough. I had spare time and gathered some of the better formulas and made a new calc sheet out of it. Still some generalizing, like a default inertia. But it seems to give quite a reasonable estimation and has much more input variables. I think most of it is correct now, but use it at your own risk.
Same link as before: http://bit.ly/16DrFdP

EDIT: Actually, it's a little bit false that it's called "stepper" now, the formulas are covering servos as well.

[MooseCat]

Same link as before: http://bit.ly/16DrFdP

I think you have an error. The forumla in cell B13 reads, "=E2*(B2/1000/2*PI())^2", but it should be "=E2*(B2/1000/2/PI())^2"
The difference is a factor of pi^4 ~= 100, in the calculation of the J-load.

With 10-15mm pitch screws and one motor on each side of the gantry you are now comfortably in nema23 land. As one should be with a mere 50kg.


On a unrelated note, kinda curious how "tight budget" and "5 axis with a tool changer" gels together.

[svenakela]

I think you have an error. The forumla in cell B13 reads, "=E2*(B2/1000/2*PI())^2", but it should be "=E2*(B2/1000/2/PI())^2"
The difference is a factor of pi^4 ~= 100, in the calculation of the J-load.

With 10-15mm pitch screws and one motor on each side of the gantry you are now comfortably in nema23 land. As one should be with a mere 50kg.


On a unrelated note, kinda curious how "tight budget" and "5 axis with a tool changer" gels together.

Thanks. I've missed the reduction too according to the Leadshine formula's. Added that as well.
It now yields "=E2*(B2/1000/(2*PI())*(C2/D2))^2". I want to keep the formulas context as they are in the PDF's so it's some what easy to follow so I'll keep the multiplication.

Gantry?! Oh no...
Tight budget means that if I can cut the costs in all there's enough money left to spend on a spindle with ATC. I can go with servo's tomorrow or start with three axis and upgrade to five axis later on, but that's not a challenge. Still, that's a whole different story.
50 kg is just a default value, X axis is at least 150.