[Vegabond]

Hello, i want to build a a-axis on my cnc, what program can i program the x-y-z and the a-axis? easy to learn


Robert.

[ger21]

What do you want to do with it? And what is your budget?

[Vegabond]

I just want to make simpel things, something like this ( [ame="http://www.youtube.com/watch?v=Mmklz5-PKXg"]YouTube- 4 axis cnc router machine[/ame] )

The budget isn`t set, i just want to see what program i can get before i set the price...


Greetings from Robert.

[BobWarfield]

The interesting thing about the two parts in the video is they basically involve mapping a part onto a cylinder. Imagine if you could unroll that cylinder and lay it flat, then an ordinary 3D CAM would work.

There are programs out there I believe that will perform that sort of mapping for you. You create a flat 3D g-code and it will roll that up onto a cylinder that you could feed a machine with 4th axis. Sorry, don't have any names or links for them.

Alternatively, if that's all you're trying to do, you could also reconnect the Y axis to drive the A and get the same result. You'd have to remap your steps per inch so it comes out right. I've seen cases where people did that too.

I'll let someone else answer for a literal CAM program.

Cheers,

BW

[dertsap]

gmax with the cnctoolkit plugin would be your best bet for that type of work
if your doing similar shapes and such without the the engraving then it would be dead simple and take minutes to hand code it

[ger21]

The video uses ArtCAM, I think. As was mentioned, there are programs that can wrap 2D g-code around a 4th axis. Google CNC Wrapper. Vectic's Aspire and the yet to be released V-Carve Pro update can wrap their toolpaths around a 4th axis as well. None of these are really 4th axis CAM packages, as they won't create 4 axis g-code from a 3D model. At the cheap end, MeshCAM can create indexed 4 axis code from an .stl. Deskproto can do 4 axis from .stl files.

After that, the price goes up rapidly for true 4 axis CAM.

[HuFlungDung]

The interesting thing about the two parts in the video is they basically involve mapping a part onto a cylinder. Imagine if you could unroll that cylinder and lay it flat, then an ordinary 3D CAM would work.

There are programs out there I believe that will perform that sort of mapping for you. You create a flat 3D g-code and it will roll that up onto a cylinder that you could feed a machine with 4th axis. Sorry, don't have any names or links for them.

Alternatively, if that's all you're trying to do, you could also reconnect the Y axis to drive the A and get the same result. You'd have to remap your steps per inch so it comes out right. I've seen cases where people did that too.

I'll let someone else answer for a literal CAM program.

Cheers,

BW

Solitary A axis moves can be scaled from Y axis output of a CAM program, if you have mapped the surface as Bob alluded to. The scale factor for Y is 360/2PI*R where R is the radius of the cylinder which is being worked on.

Alternatively, you could, I suppose, rescale the Y axis output of your rotary axis for the particular job, if you pull the Y axis motor off and install it on your rotary axis.

The 2d map of the cylinder surface is laid out in CAD so that all its features lay within a rectangle that is X x Y where X is the cylinder length, and Y is the circumference.

Arcs that exist in XY would probably best be interpolated into line segments since post processor output of G02/G03 is not going to jive with conventional XA simultaneous movement. Again, if you pull the Y axis motor and put it on the rotary, then you probably can use G02/G03......I'm not sure if there are any gotchas in there, or not

The feedrate could be more of an issue if you are trying to program A in degrees/minute. You're going to have to figure out what your controller can actually work with. There exist quite a mix of possibilities out there:
X linear inch/min.
A degrees/min
XA: what is it going to be? One or the other? both?

It would be a shame to get far into doing this only to find out your controller rules will make it really difficult for you.

[awerby]

While there are schemes for mapping the Y axis of a basically flat part onto the A axis, this is cumbersome and has a hard time coming together at the ends. If you're dealing with parts you've designed "in the round", it's not going to work at all.

DeskProto is the least expensive program I know of that will write a true 4-axis toolpath from a basically cylindrical part without trying to unroll and re-roll it. You can download it and use it free for a month before it times out; you'll find it's quite easy to learn and use. If you're still building your mill, then wait until it's operational before doing this. If you're looking to save money, DeskProto has recently decided to re-release its 4.2 version (which works fine, just lacks some advanced features and wizards) at a lower price than the flagship 5.0 version. Let me know if you're interested in that; it's not up my site yet.

Andrew Werby
www.computersculpture.com

[BurrMan]

In your video you show actually 2 methods. The first one is simple wrapping. Many programs can do this. BobCad has indexing and wrapping and can do this fairly quickly:





THe last part of your video sample shows what Ger and Hu are refering to with a possible need to have a "true 4th software".

The unwrapping method becomes "flexable" with any system that does this as you cant unwrap a true "Gaussian Curved surface".

[HuFlungDung]

While there are schemes for mapping the Y axis of a basically flat part onto the A axis, this is cumbersome and has a hard time coming together at the ends. If you're dealing with parts you've designed "in the round", it's not going to work at all.

DeskProto is the least expensive program I know of that will write a true 4-axis toolpath from a basically cylindrical part without trying to unroll and re-roll it. You can download it and use it free for a month before it times out; you'll find it's quite easy to learn and use. If you're still building your mill, then wait until it's operational before doing this. If you're looking to save money, DeskProto has recently decided to re-release its 4.2 version (which works fine, just lacks some advanced features and wizards) at a lower price than the flagship 5.0 version. Let me know if you're interested in that; it's not up my site yet.

Andrew Werby
www.computersculpture.com

I took some time to review the DeskProto videos. It looks like a pretty decent package if you use STL models and wish to do artwork machining. Makes me wish I did some

Since we don't have a Deskproto forum here, allow me to ask:

I did not notice a toolpath simulation that shows exactly what the particular choice of tool produced, or is that in there somewhere? Do you have to rely on an educated guess about what level of detail you will achieve with a given tool?

What kind of tool approaches does it offer? Is it basically plunging or are there helical entries, etc?

Is there a method to accurately pick a new datum on the partially completed part so things align after re-fixturing? What kind of precision can you attain if you try really hard? Like, how round would a perfectly round bore turn out to be? Would it be on size (if the cutter is perfect)? Sorry, I know nothing about machining stl models

[BurrMan]

For sure you would need a modeler that would allow you to control your meshing for STL.

The deskproto website says it's not for use in "Accuracy needed settings".

Although, Andrews made some small figurines that look fairly impressive.

[mohsenln90]

hello all
im new user and very happy to be here now.
best regards