Log in

No account? Create an account

Previous Entry | Next Entry

Completed X stage.

I spent much of Saturday night making CAD designs for the X stage of the the beast. A quick trip to a local hardware store supplied me with a ton of #6-32 screws and nuts, as well as #10-20 threaded rod, 1/4 inch plated rod, and various nylon and brass bushings.
Since I'd designed for #4 screws and only found #6 screws, I took an hour to tweak the cad design to fit the bigger screws. I used the laser cutter to cut 1/4 inch plywood into the necessary shapes, and then started assembling.

Laser cutting parts for the X-axis stage.

The leadscrew couples to the stepper motor with a short plastic tube. This cushions the motor from the impulse shock of turning the leadscrew in discrete steps. I'm using a leadscrew instead of a belt, because it's a lot easier to find threaded rods at an ACE hardware store, than it is to find right-sized belts.

The (more or less) completed X-axis stage has slightly more than 12 inches of travel. A larger platform will be attached to the current 5inch square surface later.

I can drive the platform with the NEMA17 motor, but I can change to a NEMA23 later, if I need to. It's rather noisy, though, so I'm looking into sound deadening solutions. I'd hoped using thicker wood would have deadened some of the resonance issues, but I appear to have been mistaken.

There were a few minor flaws to correct in my CAD design, but it all went together pretty easily the on the first try! I'm stunned I've gotten this far in only two days.


( 11 comments — Leave a comment )
Sep. 28th, 2009 05:40 pm (UTC)
Wow. You've been a busy bat. @_@
Sep. 28th, 2009 07:00 pm (UTC)
You're going to have trouble matching the speed of a belt-driven CNC bot, although you will get higher theoretical accuracy. (Don't know how the backlash will compare, though.)

That said, you'll definitely have a larger build area!
Sep. 28th, 2009 09:36 pm (UTC)
Yeah, the leadscrews make for a 20:1 speed reduction. This won't matter so much during extrusion, but it will be unfortunate during "rapid" G0 motions. Mind you, the torque reduction on the motors should allow the them to reach higher speeds, but that will only help a little. If I go to nema32 motors, I should be able to get 50 inch/minute speeds, though, as my cnc mill handles that speed just fine with a similar setup. Other options include coarser threaded leadscrews, or a chain drive. If I can find apropriate belts at an auto store or something, I'll reconsider them, though.

Backlash will likely be on par with a CupcakeCNC, due to the inherent slight looseness of the leadscrew nut. I tried a dual-nut rig, but that lead to too tight of a grip on the leadscrew, and I didn't really have any way to adjust that. This issue may just end up being a sacrifice to using off-the-shelf parts.
Sep. 29th, 2009 03:51 am (UTC)
I reduced the noise some with rubber padding on the motor mount. Final numbers on feed rate: 350+ mm/minute reliable. (~15 inches/minute)
Sep. 28th, 2009 07:05 pm (UTC)
Oooh, looking good!

I like the captive nut construction, never would have thought of that.
Sep. 29th, 2009 06:35 pm (UTC)
I stole the idea from the CupcakeCNC design. I've modified my own design since I took those pictures, to reduce the needed number of captive nuts by about half, and the remaining ones are mostly hidden from view, so it'll look better.
Sep. 28th, 2009 10:58 pm (UTC)
A rubber washer between the motor and the plywood might help with noise, though it'll reduce your accuracy somewhat.

I'd be tempted just to put the entire rig on rubber mounts inside a box with foam wedges on the walls and call it a day, but that's me ;).

Sep. 29th, 2009 06:36 pm (UTC)
I found some 1/8" rubber foam matting that I used as a motor mount gasket, and it's helped some.
Sep. 30th, 2009 03:08 am (UTC)
Filling the voids under the stage frame with something inelastic would also help damp resonances. Is there such a thing as spray foam-rubber?

Design-wise, anything that makes the boards stiffer will push up the resonant frequencies, making them easier to damp or insulate, too. I'm surprised you haven't gone for ribbed plates fabbed using a MakerBot, actually };>.

I hope at least some of this advice is useful to you.

Sep. 30th, 2009 06:23 am (UTC)
I don't actually have a Makerbot CupcakeCNC yet. That's part of the fun of this exercise. But yes, if you look at the picture, there's ribs in the underbody at five places to hold it all stiff. There is spray foam, yeah. I may end up giving that a shot soon.
Sep. 30th, 2009 06:53 am (UTC)
To clarify re. ribs: I mean more than five. This would be something approaching honeycomb panel (complexity is almost free with the Cupcake CNC, within a given raster-scanning area; I already have plots for making fractal trusses if I ever get one).

To clarify re. spray foam: I mean foam rubber, that will deform in a manner that absorbs energy. Most of the "spray foam" I've seen is rigid when it dries, and so would be closer to elastic than inelastic (it'd return most of the energy stored in deforming it, resonating rather than damping).

If they have sprayable stuff that dries squishy, though, that's extremely nifty...

On a resonance-related topic, how did the xylophone project turn out?

( 11 comments — Leave a comment )

Latest Month

February 2018


  • 17 Oct 2011, 19:18
    Clever! ^_^
  • 7 Oct 2011, 08:38
    Was each link made from two pieces, or was there fill material inside that you removed? I'm having trouble seeing how to print these in a makerbot-compatible way (though I could just be overlooking…
  • 5 Oct 2011, 22:40
    Nah. It'd just take a dissolvable support material and higher resolution. There are commercial 3d printers out there that can print ball bearing assemblies in one pass, fully assembled.
  • 5 Oct 2011, 18:46
    Nice! ^_^ I was going to be flabbergasted if you were actually able to print it as a unit, but I guess that would take antigravity.
  • 5 Oct 2011, 18:19
    I printed each piece individually and assembled them after some cleanup. I did print four segments at a time, though, unattached, in a 2x2 grid.
Powered by LiveJournal.com
Designed by Lilia Ahner