Dear CNC router experts:

I am working on the design of a very small CNC router for wax in jewelry applications.

It is a fixed gantry, moving table machine and it must have a 4th axis to
allow carving wax for rings.

However, since the machine is so small, it is not practical to mount the
rotary axis (A) on the table.

So, my question is: Am I right in assuming that, for an XYZA machine,
where A is a rotary axis ("barbecue spit") parallel to X (the gantry), I really
don't need movement in the Y axis (the table) when using the rotary axis? -
provided, of course, that the A axis is perfectly aligned under the spindle.

I think this is true and, if so, the design will be much simpler and rigid. Also,
I would only need 3 stepper drives: one for Z, another for X and the 3rd
would be "shared" by Y and A.

Nelson

Correct.

If you can lock the Spindle centered on the A-axis then you should be good.


.

I think it depends on how complex your shape is.

Very likely you will be using a small ball nose cutter. I can imagine some profiles which would require the cutter to do an undercut, which is not possib le when the the spindle and fourth axis are centerd. However if the fourth axis can move away from the center of the spindle the tool can approach from a different angle.

Thanks a lot for your replies

Actually, after posting I found a recent contribution by HuFlungDung which confirms
my assumption:

On the thread "4th axis CAM" (thread 60323 post #7) HuFlungDung says that
"Any cad system with scalable output for the Y axis could be set to output A axis code"
(see the post for details)

So, in a sense, in this situation, axis A "replaces" axis Y and you don´t even have to
use 4axis softaware - standard 3D or 2.5D software can be "fooled" to route a band
when actually "thinking" that is is carving a strip...

Thank you all

Nelson

.....So, in a sense, in this situation, axis A "replaces" axis Y and you don´t even have to
use 4axis softaware - standard 3D or 2.5D software can be "fooled" to route a band
when actually "thinking" that is is carving a strip...

Thank you all

Nelson

That is correct. On Haas machines you don't even do any 'fooling' nor do you have to replace Y with A. Haas machines have what is called 'Circle Wrap' where the machine maps the Y axis onto the A axis.

But what I said is also true; if you want to do really complex shapes on a 4th axis then you do need Y in the equation.

You might need the Y-axis since many jewelry packages use a three sided cut setup.

You mill 2 sides flat in th x-y plane and then flip again to the x and a axis.

Just something to think about.

Arquibaldo,

It is true you will limit your capabilities...a little, but I believe you can do some amazing work as you are suggesting by not moving the Y.

Dear Mike Stevenson: Thanks for the encouragement.

Dear jdr:

I am not sure I understand the details of your post: if you mean that these packages
use simultaneously XYZ and A, then I would be in trouble.

But, if you mean, as I understood, that the packages use XYZ in the 1st 2 setups
and then XZA on the 3rd setup, then I will be OK, since my design does have a Y axis.
It simply cannot use the 4 axis at the same time; you either mount the work on the table and use XYZ OR you mount the work in the chuck and use XZA.

This is because the 4th axis is mounted on the "floor" of the machine, along with the
posts for the fixed gantry and the rails for the moving table. If and when I have a
working prototype, I wil post full details.

Anyhow, the restrictions with certain software are only academic for me: I can´t afford
these expensive jewelry packages - this is a hobby, not a business; I intend to use
only free software or software that I write myself...

Nelson

Some parts will work, others wont.

An easy, no-math test to determine if you need the Y axis for a part would be to draw lines radially from the center of the A axis through the part. Make the lines as thick as your tool diameter. If the line leaves the part and crosses it again, then you cannot cut without the Y. As an example, imagine a part shaped like number "6". Center the circular part on the A axis and draw lines radially from that point. If the inside edge of the top part of the "6" is straight and along the radial line (again, the line is as thick as the tool diameter), then you can cut it on a XZA setup. If the part is curved/hooked, then the radial line will leave the circular part then re-enter at the hook. If the top tail is thinner, then when you rotate the A, the top of the tail will hit the tool, but the bottom part wont. The gap between that line and the part will be impossible to cut without the Y axis. You would have to rotate the A until that area points up toward the tool, then move the Y over to it.

Just thinking out loud now, but would it be possible to extend the capabilities of an XZA system as proposed by using a tool that has a narrow shank and a ball shaped cutter? I would think you could gain some undercut capability limited to the difference between the shank and cutting edge radius.

Bryan.