• It's time to cast your vote in the January 2025 Turning Challenge. (click here for details)
  • Conversations are now Direct Messages (click here for details)
  • Congratulations to Alan Weinberg for "Elm Burl Bowl" being selected as Turning of the Week for January 27, 2025 (click here for details)
  • Welcome new registering member. Your username must be your real First and Last name (for example: John Doe). "Screen names" and "handles" are not allowed and your registration will be deleted if you don't use your real name. Also, do not use all caps nor all lower case.

Lathe VS 4 axis CNC

Joined
Sep 16, 2019
Messages
6
Likes
4
Location
Brooklyn, New York
The discussion basically originated in the marketplace section where I mentioned that the reason to sell my lathe was a 4 axis CNC router.
CNC with rotational axis and a lathe have different advantages and disadvantages for various tasks.
Machining irregular shape or machining texture on this shape (like snake skin or a bunch of random size half spheres) on the CNC is very doable. If you can draw it, it can cut. (within mechanical limitations of the router) Physical machining is done by the router or a spindle, using various mill (ball mill, tapered mill and so on). Router moves in XYZ axis as well as the stock, slowly rotating in A axis (usual designation for rotational.) So, it is not limited to rotational axial motion of the stock like lathe and there's no need for additional jigs. You can also drill dowel holes. Basically it is very precise carving with the router.
However, it is a fairly slow process. It requires knowledge in 3d modeling chair leg.jpg twisty.jpg
as well as operating CNC router. It's a very high initial investment. The finished surface is achieved by setting small stopover on the final tool.
A lathe can output round parts significantly quicker. If you need to turn a round chair leg or anything of the sort and sand it smooth, lathe will outproduce A axis CNC by a high margin.
There are also CNC lathes. These operate like regular lathe with the exception of the numerically controlled tool feed and automatic tool changers.
 
Last edited:
Pretty cool, you’ll have to add a video of it in action once you get up and running. Was this an add on kit to an existing CNC, or did you purchase the CNC with the 4th axis?
 
It's already operational. I am in the midle of the dispute with Rhinocam (Mechsoft) which sold CAD/CAM software with problematic postprocessor but are unwilling to fix the issue unless I buy new version of the software. Yeah, cad/cam guys are that nice.
So, for these who are not familiar with CNC machining, there are usually 3 pieces of software that are involved. One is your 3d modeler. Could be Rhinoceros 5 or Fusion 360 or Autocad. That's where you build your 3d or 2d surface to be machined. Second part is the Toolpath generator. Fusion 360 has it included as part of the package and Rhinocam is a plugin for Rhino. You can specify the tool, speed and pattern in which your part will be cut. Like drilling, surfacing or roughing operation. CadCam creates G-code which is translated by a third controller program on to series of signals driving individual controler/motor combination for each axis.
Make sense?
 
Last edited:
My machine was purchassed as a kit from CNC Router parts. We used 1940s cast iron machine stands for support and welded steel frame to strengthen. It's about 2500lb now which for high speed machining is not alot. Rotational axis was added recently. Nema34 motor, controler, frame with the chuck and gearbox. I don't have anything cool to post yet.
Getting a turnkey machine of this level would cost upwords of 30k. A kit and a software is a bit over half that by the time it's all set and done.
 
It's already operational. I am in the midle of the dispute with Rhinocam (Mechsoft) which sold CAD/CAM software with problematic postprocessor but are unwilling to fix the issue unless I buy new version of the software. Yeah, cad/cam guys are that nice.
So, for these who are not familiar with CNC machining, there are usually 3 pieces of software that are involved. One is your 3d modeler. Could be Rhinoceros 5 or Fusion 360 or Autocad. That's where you build your 3d or 2d surface to be machined. Second part is the Toolpath generator. Fusion 360 has it included as part of the package and Rhinocam is a plugin for Rhino. You can specify the tool, speed and pattern in which your part will be cut. Like drilling, surfacing or roughing operation. CadCam creates G-code which is translated by a third controller program on to series of signals driving individual controler/motor combination for each axis.
Make sense?
Very impressive. Thank you for sharing. If you move to Maui, I have some ideas to collaborate...
 
It's already operational. I am in the midle of the dispute with Rhinocam (Mechsoft) which sold CAD/CAM software with problematic postprocessor but are unwilling to fix the issue unless I buy new version of the software. Yeah, cad/cam guys are that nice.
So, for these who are not familiar with CNC machining, there are usually 3 pieces of software that are involved. One is your 3d modeler. Could be Rhinoceros 5 or Fusion 360 or Autocad. That's where you build your 3d or 2d surface to be machined. Second part is the Toolpath generator. Fusion 360 has it included as part of the package and Rhinocam is a plugin for Rhino. You can specify the tool, speed and pattern in which your part will be cut. Like drilling, surfacing or roughing operation. CadCam creates G-code which is translated by a third controller program on to series of signals driving individual controler/motor combination for each axis.
Make sense?

Roman,

LinuxCNC is also a good option to Mach. And LinuxCNC is free.

That is what Bill Ooms and Dewey Garrett use.

Kind regards,
Rich
 
Roman,

LinuxCNC is also a good option to Mach. And LinuxCNC is free.

That is what Bill Ooms and Dewey Garrett use.

Kind regards,
Rich

Rich, yes LinuxCNC is a powerful control system. But it more or less comes down to being an open source system with support from enthusiasts. Centroid was my choice of CNC control over LinuxCNC because of its network of factory support people. Centroid does cost a few hundred bucks, but I have the assurance of knowledgeable support. I will likely have to pay for support if it goes beyond a simple phone or email exchange. My machines are used in business so I need that level of assurance of support.
 
I ran a 5'x10'x3' Z axis, 5 axis Motionmaster for 12 years before retiring. The one thing a CNC will never do better than a traditional lathe is surface finish. Especially if using the tip of an end mill or router bit. Carving point bits are even worse. You just don't have the cutting surface speed at the point. Same goes for either a CNC router or 3D printer. You always have a step over (no matter how small that increment was) and will always require a lot more sanding to get rid of that. With the 5 axis I could program it to cut on the radius of a ball mill, or side of a straight bit for better surface finish. But it still didn't come remotely close to the finish off a skew, or shear scrape on the edge of a bowl gouge. A CNC will not even reproduce a turning better than a high quality copy lathe with high shear angle cutters. You just can't equal a shearing cut with a rotating bit. I can see a need for CNC, but it would never replace a lathe in my opinion.
 
Kind of a horses for courses question. Seems to me that CNC excels where the quantity of work justifies the time and effort required to design and program the equipment. If I needed pallet loads of the same part CNC wins every time. Personally, I seldom make anything more often than once or twice. Being able to alter the design from piece tp piece on the fly and learning something new along the way is why I took up turning in the first place.
That said, I just finished turning 24 shaker style hanging pegs - would have appreciated being able to draw them and walk away...on the other hand, no two of the 24 are exactly the same; looks great on the wall and screams hand crafted.
 
Kind of a horses for courses question. Seems to me that CNC excels where the quantity of work justifies the time and effort required to design and program the equipment. If I needed pallet loads of the same part CNC wins every time. Personally, I seldom make anything more often than once or twice. Being able to alter the design from piece tp piece on the fly and learning something new along the way is why I took up turning in the first place.
That said, I just finished turning 24 shaker style hanging pegs - would have appreciated being able to draw them and walk away...on the other hand, no two of the 24 are exactly the same; looks great on the wall and screams hand crafted.

I don't have anything against hand turning, I've just never cared for it since I was first exposed to it in high school shop. I'd rather beat myself on the head with a ball peen hammer than hand turn. Turners get relaxation and satisfaction hand turning. I get the same thing from making a machine do what I want.

The idea that CNC excels in quantity and not one-offs. No.

The design effort between hand turning and CNC'ing should be the same. Time programming the CNC might be equivalent to the time the hand turner would spend sharpening his tools through the turning process. A major difference between the two methods is my CNC'd part will be as designed. And because of the finer cutting, less sanding needed..

Programming has gotten easier in the last ten to twenty years as the CNC controllers have improved (don't mistake what you see at Woodcraft and Rockler as modern CNC's, they're more of a step back in technology to be affordable). I can start with a scan of a pencil design sketch and software turns it into a CAD/CAM readable vector file. Design shapes can even be mathematically described and the machine will follow the path defined (macro programming). Lately I've used photogrammetry (Google on it) to reproduce turned architectural details. Once the design is in the CAD system I can evaluate it from all angles.

As far as making a run of parts to look hand crafted, not a problem. With a few key strokes on the CNC controller in seconds a part shape can be changed. With a bowl, it could be changed from round to oval (the OP mentioned a 4 axis route, CNC "turning" is better not done on a lathe, rather a 4 axis milling machine in my case).

The CNC learning curve is an issue with wannabes. With hand turning, there's a whole industry built around teaching, maybe not so different a learning curve.

Cost of CNC....... add up the cost of a top of the line wood lathe or two, turning tools, sharpening equipment, dust collector, classes, etc, etc. I bet I've got less, around $10K, in the purchased used and upgraded machine most of my woodturning is done on.
 
I don't have anything against hand turning, I've just never cared for it since I was first exposed to it in high school shop. I'd rather beat myself on the head with a ball peen hammer than hand turn. Turners get relaxation and satisfaction hand turning. I get the same thing from making a machine do what I want.

Doug - I fully understand the allure of technology - spent 40 years growing a company largely through full adoption of new, cutting edge technology. Keeping up with how that technology changes over time and making certain we were taking full advantage was a large part of my life for that time.

I’ve retired now, turning is an avocation for me. I take great pride in developing the skill to cut a flowing catenary curve from rim to base both inside and outside a piece. If I want a texture applied that I haven’t tried before, I’ll practice it until I feel comfortable with the process and tools. I realize that the time involved to create the bezier curves and textures in software may be faster and that the satisfaction and pride in making is probably the same for both of us, but I like getting my hands dirty, knocking the odd chip from my underwear when necessary and standing in shavings up to my knees.

I spent years on computers coaxing them to do what I wanted, I’d rather beat myself in the head with a ball peen hammer than go back to that.
 
Back
Top