So i have this wooden sphere, about 8" diameter, that I would like to engrave to make it look like a basketball with some additional personalized stuff on it. I asked a couple people about doing this, and basically we currently don’t have the tool or maybe not even the software capability in RDWorks to do this. The option for me right now is to use the single-axis rotary and etch along sequential circular planes (each at a different Z) along the whole sphere. This might create issues with alignment though, and there won’t really be a fully accurate focus throughout, especially given that the laser would be hitting a progressively more angled surface the closer it get to the poles.
I wanted to post this here just so I could maybe provide some brain food for anyone who has thoughts about creating a rotary tool that would allow engraving a sphere. I have an idea, but I don’t have enough experience or knowledge about the workings of the electronics/software interface that would be required. If anyone is interested, I’d love to discuss it and try to make it happen.
Have you considered making something with two rollers at a 90 degree angle? If you have ever cleaned the ball of an old mechanical mouse, imagine something like that.
The hard part would be how to make the laser not move the head along the X axis while your gadget rolls the ball. I am not sure you can do that without modifying the machine.
That’s sort of the idea I had, but like you said, I’m not sure how to interface with the machine to get the laser head in one place. If anyone here is familiar with how the rotary tool and machine interface with RDWorks, I would love some insight.
This could be fairly easily done in the woodshop on a lathe with two caveats: 1) you’d need appropriate chucking accessories (not terribly expensive nor that difficult to DIY, either); and 2) you’d have to be happy with just latitude/longitude lines, as opposed to the typical curved lines for a standard b-ball.
Given your initial post, I suspect the 2nd is a deal breaker for what you have in mind, but I thought I’d lay up (he he…) another option.
I do not believe that RDWorks is even aware of the rotary attachment. It thinks it is moving the gantry along the Y axis and the machine rotates the piece instead.
I do not believe that the Thunder has computerized control of the Z-axis, which would be necessary in order to program spherical engraving. AFAIK it has only two-axis controllers.
Bill is correct. The signals sent to the rotary are the same as the x- & y- gantry except that one of them (I think the y-) is adjusted in RDWorks based upon the diameter of your cylindrical work piece.
Conceptually in my head I think one could pull this off by using rollers to rotate the sphere while keeping the laser head stationary.
For instance say the laser is at (above actually) the North pole of the work piece. And then have 4 rollers along the equator spaced 90 degrees apart holding the piece and rotating the piece. The South pole could be a low friction curve that prevents the piece from falling.
The four rollers along the hemispheres would likely have to be designed with 2 degrees of freedom: 1) to roll the work piece (ie. rotate along the axis of the roller), and 2) rotate while the other two rollers are rotating the work piece (ie. rotate out of the plane of the equator).
I’m thinking the s/w might be fairly sophisticated but certainly doable.
This is pretty much the exact idea I had in my head, only difference is that I would put some spring-loaded forced on the four arms that hold the four rollers so that 1) there’s enough grip on the sphere and 2) the tool can be used for spheres of various sizes. I’m imagining these “rollers” as rubber spheres themselves, like the old-timey computer mouse wheels. And for the “south pole,” maybe just a ball-bearing base that allows the sphere to roll with low friction.
With the above design, the laser head would be essentially stationary and would not need a dynamic focus or computerized control of the Z-axis. The contraption would rotate the sphere and do all the work.
Given the limitation of RDWorks, I think the laser would still have to do it in “slices” along the sphere, with this tool simply allowing for precise and minute rotation of the y-axis after each slice has been engraved through the x-axis rotation.
“Given the limitation of RDWorks, I think the laser would still have to do it in “slices” along the sphere, with this tool simply allowing for precise and minute rotation of the y-axis after each slice has been engraved through the x-axis rotation.”
Yep. This methodology would work for rastering but not for vector artwork.
If one had a square image that they import straight into RDWorks, there would be overlap in the corners/edges when trying to use this methodology.
I guess another potential workflow for this methodology would be to use another piece of software prior to RDWorks import and perform these transforms/steps:
Wrap image around globe of correct size.
Project desired view of the globe on to a flat plane.
Save projected image to file.
Can’t honestly say if this way would be any better. Just a thought.
Would be really awesome to be able to pull polar coordinates from a file and be able to perform vector artwork.
Having 2nd thoughts … thinking leave north pole as is (ie. laser location). Get rid of the equatorial rollers and replace them with 3 balls (setup like a ball bearing) spaced 120 degrees apart from one another and located at 45 degrees N latitude. And then place a single roller (with 2 degrees of freedom) at the South pole.
Lasering at the North pole would be crucial for avoiding focus issues.
I don’t know how you could do a roller with 2 degrees of freedom with full rotation on both while moving them simultaneously (I might be missing something). Seems like it would be easier to have 4 rollers in a “cross” formation at the bottom with 2 controlling each rotation direction? (if that makes sense)
You would definitely be translating from another coordinate system to Cartesian just to run the motors and not worrying about what it “looks like” in the controller software (RD Works etc.)
“I don’t know how you could do a roller with 2 degrees of freedom with full rotation on both while moving them simultaneously (I might be missing something).”
Make no mistake, I might be missing something!
But this was what first came to my mind … the roller is actually two rollers, one of which is motorized, and the two rollers in tandem can be rotated by another motor. Apologies for the crappy drawing.
Ah, I think I get what you’re saying. The motor at the bottom right of the image would be facing up?
I was thinking something like the following. Imagine two of these 90 degrees apart and one of the axes offset with some gears so they don’t hit each other. The sphere would sit between the two rollers (cylinders here on the axis). The cube is a motor.
