Stepper Motor Torque

[donek]

I'm building a cnc bender for bending ski and snowboard edges. This device consists of 3 rollers. One is fixed free spinning (essentially a bearing), one is driven by a stepper motor and is used to feed the edge, the other is free spinning but moves on an eccentric driven by a stepper motor. I am using an old shopbot control box with a 28v power supply.

I initially powered the driven roller with an oriental motor PK 299-01AA


holding torque: 6.2N-m
step angle: 1.8 deg

It was unable to supply the necessary torque to feed the edge through the device. I am looking at replacing it with a geared stepper PK296A2A-SG10


holding torque: 7N-m
step angle:0.18deg
gear ratio: 10:1

I might also consider a PK296A2A-SG18


holding torque: 9N-m
step angle: 0.1deg
gear ratio: 18:1

The motor that is connected to these devices is a PK296


holding torque: 3.1N-m
step angle 1.8deg

My desired speed on this motor is between 10 and 20 rpm.

In my world a 10:1 gear reduction should result in 10 times the torque, but the specs on the oriental site indicate slightly over half. Am I thinking in the wrong terms or missing something? Will the 10:1 gear ratio give me 10 times the torque at these speeds or do steppers work in a manner that is completely different from conventional gear motors?

I apologize if this question has been asked and answered before. I was unable to find a similar discussion in my searches.

[richards]

Your math is correct. The reduced torque (compared to what you expected) is caused by the gearbox that Oriental Motor uses. The gears cannot handle the torque developed, so the gearboxes are under-rated. Other motors/gearboxes are available that can handle more torque but their price is substantially higher.

If you built a two-stage belt-drive, you could get the expected torque.

Remember that the rated torque on a stepper motor is usually its "holding torque", or the amount of torque available when the motor is stopped and full current is applied to the motor's coils (i.e. excessive heat). Most stepper drivers have automatic current reduction that cuts the current to the motor by half or more when the motor has stopped to reduce heat.

If you study the torque curves published on the Oriental Motor web site, you'll see how much torque to expect at various speeds. Pay attention to the voltage that they use and realize that they also specify using a large heat-sink on the motor (which should tell you that the motor is going to get really, really hot when run under the conditions that give you the published torque).

Finally, realize that the "duty cycle" of a stepper motor is not 100%. Even with a large heat sink, running a motor that pulls the maximum current for extended periods of time will fry the windings.

None of that is normally a concern on a CNC machine. Motors are usually lightly loaded (compared to the Oriental Motor test conditions).

If you visit the www.geckodrive.com (http://www.geckodrive.com) website, click on the "support" button where you'll find an article on stepper motor basics. The article clearly shows the relationship between torque, speed, voltage and heat.

[Brady Watson]

You'd think it was 10x the torque, but it isn't because the motor on the gearbox variant is actually less powerful than a direct drive. Plus, the direct drive is wired bipolar parallel, where the geared motor is bipolar serial...check the wiring to be sure.

You are not getting max torque out of those motors with that drive system. The power supply voltage and current will not give the motors all they can deliver. At the cost of speed you can increase the torque by adding your own gear reducer to your motor - timing belt, screw etc.

-B

[donek]

Thanks guys. I'm having a tough time finding timing belt gears that will produce the gear reduction I'm looking for without a double reduction. I did look at chain drives. The motor will only run in one direction , so backlash is not an issue. Can you think of any reason to avoid a chain drive?

[Brady Watson]

Chain is sloppy...expect some small positional error - but nothing major if you are only going in one direction. You're going to max out around 4:1 with any single stage timing belt setup.

-B

[richards]

Brady brought up a very good point. Stepper motors need stepper drivers to make them work. The PK296A2A-SGxx motor really needs a Geckodrive stepper driver to get full potential out of the motor and it needs to be wired half-coil.

The 4G upgrade from Shopbot uses those stepper drivers. You may need to change current limit resistors, some wiring, and change power supplies to take full advantage of that upgrade. (A 35-volt DC LINEAR power supply with a 12,000 uF or larger filter capacitor is ideal for the PK296A2A-SGxx motor/gearbox. Some people use switching power supplies, but stepper motors and switching power supplies don't always play together nicely because a stepper motor continuously varies the current that it draws and switching power supplies like to see a fairly constant current draw.)

Yes, you'll need to use a two-stage belt-drive to get a ratio higher than 4:1 (18-tooth:72-tooth). A 2-stage belt drive would give you up to 16:1. Using chains and sprockets is outside my range of experience.

[donek]

Sounds like I need a conversation with someone who has a good grasp on these stepper motors. Too many variables to sort out in my head from reading disjointed web postings.

I went hunting in the junk today. Turns out the old Hitachi rewsaw motor I kept has a 10:1 gear reducer built into it. With a little machining, I've got a 10:1 gear box. I'll cobble something together tomorrow to see if it's going to work before I do the final version of this thing.

I'm working with parts and pieces that were on the shelf on this thing. I obviously could have done better if I'd ordered everything to fit.

[donek]

I've almost finished my project and thought I would share a video here. I used my old PR controller to run this thing along with the motors previously discussed. With a 9 to 1 gear box, I got the torque I was looking for. I'm sure there is a lot of room to improve the device, but it should work for this next season. Perhaps I'll revisit things in another year and refine it more. I owe a lot of thanks to the guys at shopbot and those who responded on the forum here. Without the help and guidance I received, I'd probably still be floundering.

http://youtu.be/kMtEVDG4K54

[curtiss]

My first skis just had "maple edges" which sort of "lost their edge" after the first season.

Uphill transportation was often a bit crowded in the mid-west...