[WarrenW]

I need to couple a nemo 23 267 oz/in stepper to the Acme 1/2-10 threaded rod. I have seen where people use rubber hose and clamps, some say to machine down the rod, others say a love coupler.

I don't have access to machine the rod any. Whats the best way otherwise?

On the other end of the rod I was going to let the rod rest in a bearing that is held in place by wooden blocks. Is this all I need to do? Its just a normal 1/2 inch bearing that the rod fits inside perfectly.

Warren

[widgitmaster]

The rubber hose and clamps will work if all you do is wood, but what about a bearing?

What about endplay? The force from the stepper will want to move the screw back and forth!

Eric

[jeffs555]

Warren,
Like Eric said, you will need at least one bearing at the motor end of the shaft, to take both the thrust and radial loads off the bearings in the stepper motor. You can put one 1/2" bearing at each end of the rod, and then put nuts on each end of the rod to preload the bearings and tension the rod.

For driving the rod from a NEMA 23 motor, you can get some 1/2" inside diameter reinforced rubber hose and hose clamps. If your motor has a 1/4" shaft, you can get a setscrew type shaft collar that has an inside diameter of 1/4" and and outside diameter of 1/2" to make the motor shaft fit in the 1/2 id hose.

Jeff

[WarrenW]

Thanks! I can put a 1/2" bearing on each end of the shaft. When you say to put nuts on each end of the rod to preload the bearings and tension the rod, what do you mean by this? I just thought you meant to hold the bearing in place on the rod.

[fyffe555]

The motor should not carry any load axially or radially, the motors bearings are not intended to carry the thrust of the leadscrew. You need to securely mount the leadscrew first and then attach the nema23 motor after. Usually you need to locate one end of the screw and let the other float axially, or arrange the bearings to put the screw under tension. Tension will help reduce whip. Don't put the screw under compression because it will make it whip at lower speeds.

To join the motor to the screw with significantly different sized shafts on a home built router then a spider coupling would be the easiest and cheapest way. Mcmaster page 1041 has them. Three parts two hubs, one on each shaft sized for the shaft and a rubber spider in between provide the join.

There have been comments that the rubber spider coupler could allow backlash but if you read the specs the actual max value of missalignment under load is 1 degree. On a 1/2-10 leadscrew that's half a thou and thats under heavy load. You'll have more than that in the rest of the machine.

If you can get the ends of the screws turned to produce a flat shaft then that would help both bearing mounting and coupling mounting and make a stiffer machine.

EDIT: sorry Jeff, just repeated waht you said...

[jeffs555]

Warren,
This simple drawing shows what I mean. Starting from the left, you have a nut, then a bearing, then a bearing support, then the other bearing support, then bearing, then nut. The nuts force the bearings against the supports, taking out any axial play in the bearings, and also tensions the screw to reduce whip.

Jeff

[WarrenW]

Thanks for the drawing Jeff! Thats exactly what I was thinking of doing. Just forgot I needed a bearing at the motor end. Otherwise like you and fyffe555 said, the motor bearings would not support it well enough. I'm cutting the gantry pieces now so I'm on my way!

Warren

[InspirationTool]

Warren,
This simple drawing shows what I mean. Starting from the left, you have a nut, then a bearing, then a bearing support, then the other bearing support, then bearing, then nut. The nuts force the bearings against the supports, taking out any axial play in the bearings, and also tensions the screw to reduce whip.

Jeff

Dumb question here, but would it work okay/be better if you had the thrust bearing on one end and a radial bearing on the other end? That way if the rod got warm, it wouldn/t get loose? Or is this just not that big of an issue?

Could you put a thrust bearing on one end and use the stepper for the radial bearing on the other end?


-Jeff (a different one)

[jeffs555]

Unless the screw is very short like on the Z axis, you will want radial support on both ends. For high loads, you would normally use angular contact bearings, but they are expensive. For cutting wood and light aluminum, deep groove radial bearings are OK. I don't know how much of a problem elongation and loosness from heating of the screw is. Many designs use back to back angular contact bearings on the motor end, and radial bearings on the opposite end. The bearings in the motor should only support the motor shaft and nothing else.

[ger21]

Go to www.nookind.com and look at their acme catalog. Lots of good info there.

[jeffs555]

Jeff,
Your question about heating in the screw got me to thinking, so I looked up the thermal expansion rate of steel. It is .000012 x length x temp change in deg C. For a long screw this could be significant. For a 42" long screw that is about .0005" change in length for each deg C. I don't know how much a screw would heat up from the friction of a delrin nut.

Jeff

[spalm]

Jeff(s),
I am using double radial bearings on the motor end and single radial on the far end and it seems to work fine for me. But just for the record, do thrust bearings somehow compensate for any expansion? Do large production machines need to concern themselves with this? Or do ballscrews generate less friction which means less heat which means less expansion?

Not really a concern for me, just want to know.

Thanks,
Steve

[chuckknigh]

Some bell washers on the ends, properly tensioned, could easily take up that amount of expnsion or contraction.

It would however introduce some play into the system...so use them sparingly.

-- Chuck Knight

[ahren]

Spalm's approach is the best way to compensate for thermal expansion -- double bearings on one end, and nuts on either side, coupled with a "floating" bearing at the other end of the screw. That way, the screw is free to grow (except for the ~1" or so in between the double bearings). This is how we do it on our ballscrew-based pick and place robots here. As others have mentioned, this method doesn't tension the screw to reduce whip, so for a hobby machine, Jeff's approach may be better. There's nothing special about thrust bearings to compensate for thermal expansion -- they're just made to act as spinning "spacers" that can take an axial load.

As for a cheap way of coupling to your motor, I second the choice of a lovejoy coupling, but I would go with www.use-enco.com. I got mine from there, and all three pieces of the coupling cost me a total of $5.15! Plus, on orders over $50, you can get free shipping. That's tough to beat.

[WarrenW]

Thats what I'm gonna do. Thanks! And I love ordering from use-enco.com because I get the products the next day with ground shipping. They are out of Atlanta I think and I'm just below them in Savannah.

[DennisCNC]

Don't hurt the poor fish :cheers: