I'm still lost making a choice in machines.

[Andrew96]

I hope this post is in a correct forum.
I don't know where to start anymore. After a couple of weeks of reading around in circles..I need to start again. Hopefully, you can help me start to choose a machine again. The process might be like this. You have a part you'd like to make, choose a machine that is sized properly etc, then look at the tool path to see if the machine will do what you want it to do. 'What a CNC router can do', seems to keep coming back to the code it can run, and if you can create it. Hence why I'm here in the CAM forum, the main control room.

I'm a furniture maker. I'll assume some of you might be so you know you can make tools do things they may not be really designed for with some creativity and fixtures, wood is soft and rather forgiving. You know...cut a cove on a table saw etc. I'm not looking to make a CNC router do everything, I have lots of other machines for certain jobs. I want it to do what I can't, well, can't do easily.

I'm looking at buying a CNC router to help me contour chair seat bases with some repeatability. Of course, that led to the ton of things I could also do with a machine like this..but now I'm lost again. What I'm stuck on now is do I need a true 3 axis setup? Or will 2.5 axis code work well enough for me? It might be nice to do some trick embellishments in the future with some carvings or what not..but for now, the driving force for the machine is some perimeter routing (somewhat easy 2D stuff) and contouring.

As I understand it.. a machine that is running WinCNC has the ability to move 6 axis simultaneously. The one I'm looking at has 3 axis all using steppers. Is it all about the code you feed it? I can't imagine a 2.5 axis being much good...unless you can move say X and Z, or, Y and Z at the same time. People say X and Y at the same time though. If 2.5 really means say X + Z at the same time, you could then create a contour the same way I would with a copy machine.

If you're still with me, can I create a seat contour using software that creates 2.5axis code? Or do I need to step up in the software cost and go with 3D software that is 3 axis control?
Can you think of a part that can't be made with a 2.5 axis router, but can with a full 3 axis? I'd like to pick out the difference, and can't imagine it right now.

Any help would be appreciated.

[ger21]

Is it all about the code you feed it?

Yes. There's really no such thing as a 2.5D router, unless the Z axis is not CNC controlled.

I can't imagine a 2.5 axis being much good...unless you can move say X and Z, or, Y and Z at the same time.

You just said the driving force is perimeter routing, which would be X and Y at the same time. If you're not doing carvings, the majority of what you do will fall under your "not much good" label.

When you say 2.5D, you're talking about software. And 2.5D software is all about X and Y. Can you jump through some hoops and cheat a little. Probably, but I'm not going to go there now, because you'd really need to know what your doing.

If you're still with me, can I create a seat contour using software that creates 2.5axis code?

No. Technically, it is possible, but realistically, no.

Think of 2.5D as cutting parts out of a sheet, and 3D as shaping parts. I occasionally make mouldings using 2.5D CAM, but it can be slow, tedious process. Carving a seat would be an order of magnitude harder.

2.5D CAM programs import 2D line drawings. in order to cut 3D shapes with a 2.5D CAM program, you have two complex, and important steps.
One. You need to actually draw all of your toolpaths in your 2D CAD program.
Two. You need to import that 2D drawing into a CAM program, and tell it how deep to cut each toolpath you drew, one at a time. And keep in mind that these toolpaths are X and Y only.




One thing your leaving out, which is probably the most difficult part, is that you need to create a 3D CAD model of your carved seat. Can you do that? It's not as simple as buying a CAD program and just doing it. Modeling complex shapes can take a lot of skill and experience to do.
But, once you have you're CAD model, it's a very simple process to load it into a CAM program, and create the code to carve it. And it doesn't have to be terribly expensive. programs like MeshCAM and Vectrics Cut3D can do it with no problems.
Be aware that it won't be a fast process. If you want a finished part to come off the machine, expect it to take up to an hour or more. A lot of factors can come into play, though, so time can vary by a large amount.

Can you think of a part that can't be made with a 2.5 axis router, but can with a full 3 axis?

Anything that's not flat.

Bottom Line.
Unless you're very, very skilled, at both the CAD and CAM programs you're using, then you need a 3 axis CAM program.

[Andrew96]

Ger21, Thanks for the quick and detailed response. I know you must have read my post quickly since the driving force for a CNC router is to perform contouring, not just perimeter routing as you thought. That's a nice extra. However, you have managed to answer everything else. Thanks

Without going into great detail and creating a huge post, I did skip over quite a few things, so it does sound like I'm just some sort of wood hacker, looking to get into a machine that can create complex shapes for me, without much work. In my previous career I spent some time integrating robots into work centers running various programs via PLCs that I either wrote, or edited. I've seen a bit of code. There are lots of things a 2.5Axis system could do without getting very fancy with the program, You pretty much told me the route to take is 3D. Anything other than flat = 3D. Perfect, thanks.

As you might expect then, I'm onto the next step, going from my drawings to some CAM program to generate code, once it's got some depth to it. One of the reasons I sought your opinion on the code axis issue was to make sure I'm after the proper software to actually create the code, as you state, a rather important part of the process. The software comes in so many varieties it's hard to narrow a family down.

I seem to think I'm pretty good with ACAD. I've been using it for many years, created countless drawings. All of the parts I currently produce are already drawn in 2D using AutoCad. I've never had a reason to upgrade anything to move to 3D via solid works or countless others since I run NC machines for my metal working, or manual machines for my wood. I certainly know how much work can go into creating a complex shape like a seat contour. Currently I cut them by hand, from a drawing and my repeatability is via templates across a locating fixture. I believe the CAM software I've looked at can all accept Dxf files to then create the profile but I'm only just getting into that portion of the project.
Drawing a seat contour by hand would be tough I agree. However, one machine I was looking at has a digitizing head. I have not used anything like that but from some quick reading, I think it could create a file in some format (Polyline, point cloud in ACAD) that I could modify either with ACAD I own, or as part of a package I purchase. Obviously I've got a lot more research to do but I've got to narrow the search down a bit. I'm not going to take a year to get into a machine, but I won't go into one blindly either.
So far I'm starting with a scenario:

1 - Drawing into ACAD
2 - Output to 3D software to input arcs (compatible 3D software)
3 - Output to Code generator (could be integrated with above)
4 - Output to winCNC to follow code.

For this machine, 1 and 4 are known. The CAD, CAM or CADCAM package are the unknowns now. Might you be able to suggest some of the more popular ones that might do what I'm working towards? If 100s can do what I'm after can you suggest another method to narrow the search?
I have not looked at meshcam yet.

I do understand the length of a run it might take to create a beautifully smooth finished part. An hour...hmm, I can't do it by hand that fast, so sounds pretty fast to me! I'm thinking a balance between a roughed out shape and too much hand work. Some hand work is perfectly fine, there are not many steps that everything comes out with a perfect finish.

[ger21]

2 - Output to 3D software to input arcs

Don't understand what you're saying here??


1) AutoCAD >> This is your CAD software. Can you create your seat as a 3D model in AutoCAD? Or is that what you meant above?

You need a 3D model, whether done in AutoCAD or another 3D modeler.

2) CAM >> This should load you're 3D model, create toolpaths, and export g code.

3) Load g-code into WinCNC to run the machine.

However, one machine I was looking at has a digitizing head. I have not used anything like that but from some quick reading, I think it could create a file in some format (Polyline, point cloud in ACAD) that I could modify either with ACAD I own, or as part of a package I purchase.

The typical procedure when digitizing is to use a program to convert the point cloud into an .stl file. Then load the .stl (which is a 3D model) into your CAM program (#2 above). Editing point clouds is not for the faint of heart.
I wrote a macro that reads a point cloud and creates a mesh in AutoCAD, but it only works for a very specific format of point cloud, and wouldn't be of any use to you. There are commercially available add ons for AutoCAD to import point clouds, but they aren't cheap.

As for recommendations. Personally, I'd model it in AutoCAD, and use MeshCAM. That's what I use. If you can't do it in AutoCAD, maybe try Rhino. But I doubt it will be much easier. It's a very difficult thing to model. Any program you try will have a very large learning curve, and there's a lot more to it than just knowing how to use the software.

MeshCAM will do what you want, and it's inexpensive. Cut3D is similar. You might want to look at CAMBAm. Never tried it, but people like it.

Anything else gets a lot more expensive in a hurry.