r,theta,z VS x,y,z designs.

[cnczane]

Offered for consideration for those who have not yet plopped down
good money for parts.

An alternative to building conventional, XYZ-based moving-gantry or moving-table designs would be to use a rotating table.

Such a machine would use cylindrical (R, Theta, Z) coordinates, and would be attractive for featuring a fixed gantry like the moving table and a smaller footprint like the moving gantry. A distinct advantage is that one of the axes is directly driven by a stepper.

The $64,000 question: Can a rotating table be made stoutly, and cheaply, and precisely enough to make the approach feasible?

Another way to put the question: Is there a relatively hefty and rotationally precise "thing" out there that can be had cheaply enough to serve? Or Could one be made at least as cheaply as the XYZ approaches?

I may have answered my question ("If it's precise, it ain't cheap.").

(I think I've been eating too many leafy vegetables...)

[jimbo]

Are you talking about a design similar to this?

http://freeandeasy.sourceforge.net/rotproto.php

Software to drive the machine will be the biggest obstacle.

If this is not what you're referring to, do you have a link to an illustration that describes it?

[NeoMoses]

Dan, I've had the same idea, but mine was inspired by radial drill presses. If you're looking to use a rotary table, however, I think they're plenty stiff for wood and metal milling.

I, too, think the problem will be software, not hardware. If you can find CNC controll software that will control it and CAM software for it, I'd say give it a go.

[NeoMoses]

Originally posted by jimbo Are you talking about a design similar to this? http://freeandeasy.sourceforge.net/rotproto.php Software to drive the machine will be the biggest obstacle. If this is not what you're referring to, do you have a link to an illustration that describes it?

That's very interesting, jimbo. That's actually a Theta, Theta, Z machine (well, the Z axis isn't actually implemented yet). I've never seen a machine quite like that, with 2 rotary axes to provide X and Y motion. Thanks for the link!

[oz9ny]

Originally posted by NeoMoses That's very interesting, jimbo. That's actually a Theta, Theta, Z machine (well, the Z axis isn't actually implemented yet). I've never seen a machine quite like that, with 2 rotary axes to provide X and Y motion. Thanks for the link!

Hi

Elektor Magazine has a PCB drilling/milling machine like that in 2001 named TanBo

www.radixgmbh.de makes a kit.

Linux EMC can drive the machine

/niels

[ballendo]

Hello,

For drilling or pick and place this theta-theta (or theta-linear) is a nice solution. BUT...

For milling it comes fraught with problems. The most of which is axis velocity requirements for real world machining.

Picture the machine making a straight line through the rotation axis of the table. In the centerof the table most linear moves will require one axis to be moving as fast as it cam, whilst the other is moving at its slowest,or stopped. And even with that, there will be inevitable dwells for many move near the center of the table travel...

As mentioned above, not a problem for a drill or pick and place. But for a mill???

Hope this helps,

Ballendo

P.S. thanks for the Tanbo link. I hadn't seen that as Elektor has been dropped by my local bookstore. And also for the other link. I posted on CCED awhile back about this machine, but had lost the lURL.

[cnczane]

jimbo,

I was contemplating something simpler than the freeand easy. A good way to put it would be: "If your workpiece lay on a lazy Susan, what kind of a CNC router could you build?"

oz9ny,

The Radix has the essentials. Make it about 24" in diameter and give it about 6" of Z travel, and that's more like it.

NeoMoses,

Software was the least of my worries, I thought, because after all, the coordinates are a mere sine/cosine away from each other.

ballendo,

I don't follow (but luckily I've read your other posts so I know better than to bet against you... ). You are saying it's goiing to be hard to cut, say, even a small square around the center of rotation. I sense you've found the weakness, but "What's a dwell?"

I'm grateful to you for bringing up the velocities of the axes. Anyone had calculus who feels up to it?

("This guy wants to do CNC, but he doesn't know 'What's a dwell?'")

[ballendo]

Dan,

Dwell in CNC means a pause.

So what I'm saying is that when you have a rotary table making linear moves; sometimes the rotation needs to happen VERY quickly to keep the line straight. And as the rotary table of amachine BASED on one is going to be large, this required quick rotation is not going to be easy to achieve...

Especially when you consider that there are times when the rotation needs to be many degrees, while the "other" axis moves a step or two! And if we're machining at anything like "normal" speeds, it just "aint" gonna happen...

Hope this helps,

Ballendo

P.S. So during the dwell the cutter is not moving at cutting speed and therefore cuts deeper into the material at the points where the rotation is greater than can be accomplished in relation to the 2nd axis and feedrate.

But for drilling it's fine, because the dwells happen "above" the work.

[cnczane]

Thanks! I see it!

I knew the circular system would have to move in weird ways to cut rectilinear pieces, but I was thinking "that's only a computer problem, and it can easily keep up."

I forgot about the physics. Moving a massive table (and work thereon) at high accelerations means high (read, expensive) forces.

I have new respect for my XYZ machine.