[Winter]

Hello All.

I'm working on a custom design large cnc router and looking for your comments o n my choices.

Some basics
2800mm x 1450mm
ethernet smoothstepper
SBR20 rails and blocks (Chai's) on all axes.

X axis:

2x 2800mm RM2505 ballscrew
2x Nema 23 387oz Stepper, G251s, slaved in mach3
Rotating nut design
48v PSU for two steppers

Y axis:
1x 1450mm RM2005 ballscrew
1x nema23 570oz stepper. G201X
48v PSU shared with Z

Z axis
1x 300mm RM1605 ballscrew
1x Nema 23 387oz stepper. G251.

2.2kW water cooled china-spindle and vfd.



The hard part will be the rotating nut design I suspect.

I've attached some renders of my design so far..
The renders are missing a bit of detail (bolts and nuts, aluminium angles etc) so it might not be clear how it bolts together, but you should get a rough idea.

The rotating nut design is basically a timing pulley bolted to the ballnut, sandwiched between two aluminium plates with three-piece thrust bearings between the plate and the nut. -see photos for more info.

Gantry is made of two 100x50x6mm C channels bolted one ontop of the other, with the rails bolted to the flat face.


I can post a solidworks e-drawings file if anyone is interested in looking at it in more detail.

Comments please... I already have all the parts listed above, so there is no turning back now!

[Phife]

Just a few suggestions, I would see about redesigning that rotating nut, having it cantilevered out like you do and with the forces it may see could cause you some problems. Could you post up on here some clearer pictures that arent so shiny so its more clear how it all goes together?

Also, C channel is not known for being very flat and straight, you may have issues with the surfaces not being square.

Also, you will want to reorient your SBR rails on your gantry, they way you have them now, the Z axis will want to pull off the gantry beam pulling the open bearings from they rail, this is the direction these rails and bearings are weakest in.

Try putting the rails on the top and bottom of the Gantry and the rails on the Z on either side of the Z plate instead of the front.

[Winter]

Thanks for those suggestions.

Here is a slightly less shiney render of the rotating nut, as well as a noted drawing which might explain it a bit better.

I had thought that I had got the driving mechanism much closer to the gantry ( and thus cutting tool) than many designs - didn't think I would have problems with the small cantilever there. What problems do you suggest I might encounter?

That rotating nut assembly is 10mm plate ali - surely there wouldn't be measurable flex in that over 250mm? I could always add angle on the back to strengthen if required..

[harryn]

I sure wish I know how to use CAD, those drawings are always cool.

I can't tell for sure, but it looks like your frame is not nearly stiff enough for that size setup. A nearly 3 meter long section will need something more like a 300mm deep rectangular tube to support it with low flex, and the gantry will need close to 200mm x 100mm tube. ( My gantry is nearly the same length, so I recently ran the numbers) 6mm is a good thickness.

The legs as shown will wobble from the forces of the gantry moving back and forth. We all get fooled in our designs by thinking of our kitchen tables, but the power is really dramatically higher than we are used to.

As Phife points out, C channel looks stiffer than it really is in practice. Consider to go directly to square or rectangular tube instead for everything.

You will find out pretty fast that the desired tubes are pretty heavy. For that reason, some people (me) are building the table closer to the ground.

If you can, design the dust and noise control enclosure in now while it is relatively easy.

[dodger889]

You might look for a extruded pieces for the gantry cross support for a truer straightness. And it would be less mass to move. This is a personal thought here I would run a smaller drive pulley for the ball screws.

This is added afterwards and some thought. I would run the rail on top and bottom of the gantry support less torquing affect on it. This is how I would do it myself.