So I decided I should just do a thread on this little project that got way out of control instead of hijacking Shota's thread. This will take several posts to get a long story up here.
I’d like to preface this with a little note about me. I have an electronics background, a machining background, and I have been using AutoCAD professionally for years. I definitely don’t recommend doing this but it’s the journey I went on. I did it as much just for the gratification of pulling it off, as because I needed it. So I wouldn’t recommend doing this but I’ve enjoyed it. One of my main goals was to be able to produce a windless in an easy fashion. So how do you make a $800 windless? well follow along.
I'm not going into software discussions in this thread. I will say that I use Actcad which is a inexpensive version of Autocad, Fusion 360, Mudbox, Meshlab, and Deskproto. This machine uses a 4 axis GRBL controller. It's been discussed in several other threads on this site for both milling and 3D printers. The machine is the easy part. Learning and using the software is the big challenge. At any rate that isn't what this thread is for so I'm not going into any more of that.
A few years back I bought a run of the mill 3018 CNC machine because I wanted to try it out. I bought a Fox
3018Pro-SE. Like this.
At the time I was messing around with radio controlled airplanes and saw where a guy cut parts from carbon fiber sheet to create a wing folding mechanism for an F14. Long story, but bottom line is he was successful so I thought I would give it a try. He had talked about the power of the spindle so I bought an upgraded 300 Watt spindle at the same time. That was a good decision.
When I got it is was a pretty decent looking package but the entire z axis assembly that holds the spindle motor was a piece of printed plastic. The supports on the sides were some type of acetal material. The 10mm diameter rods were very flexible. Basically to make a long story short it was really easy to flex the tip where the tool mounted. Easily a 1/4” of play so there was no way this was going to do much in the way of accuracy at any reasonable cutting rate.
I had a 7x14 mini lathe and a manual benchtop mini-mill (variant of the Sieg X2) for quite a few years. I successfully converted the mill to CNC to make shock towers and chassis for RC Cars. I learned a lot during that project and the years of machining in the garage. One of the biggest things I learned is the practicality of using a mill to cut a particular material has a lot to do with the rigidity of the machine. The sieg would cut stainless steel pretty well but you had to go very slow. Once again to not go into too much detail, but if the machine starts to flex, the tools chatter and or break, the cuts aren’t accurate, the motors will lose steps, etc.. This forces you to slow down. There is a point at which you have to go so slow that it becomes impractical. I saw a beautiful relief carving done on 1 of these 3018 machines. It was very impressive but the guy that did it said it took 12 hours to machine, OUCH!.
At this point I knew this machine wasn’t going to do what I was hoping for. Not being one to give up particularly easily. I decided to scrap the basic frame and z axis of the fox alien. I basically stripped it for screws and miscellaneous hardware. I looked around and at that time they had just come out with the 3018 Pro Max. It is all aluminum, has 12mm rods for the bearings, 10mm aluminum plates for the sides and front and back and twice the size of extrusions for the rest of the frame. It also has more powerful stepper motors. They still sell this machine on Amazon for $277 USD. It's reasonably robust and is truly capable of relief carving at an acceptable level. Here’s a photo.
I used my 3D printer to make mounts for the limit switches that came with the Fox and put limit switched on the Pro Max. Limit switches are something you can get by without but you will repeatedly ram the machine into the stops and this really isn’t good for the motors and electronics. I’ve somehow managed to never damage a motor but I’ve sure strained the heck out of them. Throw in the fact that they can be used to automatically home the machine and personally I wouldn’t run a machine without them. So here is how the 3018 looked at that point.
So time went by and I never did anything. Summer happened, and I wasn’t flying planes due to camping trips etc. This machine sat up on top of a cabinet in the garage for about a year and a half and I never really did anything with it. Last fall I was asked to help some friends with a model ship project. I actually spent a lot of time thinking about it before I finally decided to go ahead and do it. After I decided I would do it, I was asked to do something I’ve never done, and that was create 3d sculptures for the ship.
I definitely had my doubts about pulling that off, but figured I might as well try. After a couple weeks of trying I was literally preparing to send an e-mail explaining that it is way beyond my abilities. When I had thought about a different approach. I had a breakthrough and figured out a way someone with crazy tech skills; but weak artistic skills could do sculpting. That is a whole other very long story. The bottom line is. I did eventually manage to figure out a way to do what I consider passable relief carvings at least.
That got me to wondering If I could reproduce them on the 3018. My first try was with a 1/16” ball end mill and here was the result. Not great detail but it actually worked. A 1/16" endmill at this scale is a roughing tool, way to big to do much detail.
Then I did a little more work on toolpath and speeds and feeds and produced this carving. I showed these in Shota’s 3018 thread. The video is the top piece being cut in hard maple and the lower is Alaskan Yellow Cedar. These are finished with a 1/32" tapered ball end mill.
I was encouraged by this test to the point I really wanted to try and set this up with a rotary axis. I looked online for a controller that would work for a 4th axis and it turns out they do make one, so I hunted around for a 4th axis that would fit the machine. The smallest 4th axis I could find had the center of the 4th axis 44mm above the table. The Pro max had 50mm of travel. So I did some measurements and if the spindle was all the way up in the mount. It could lift the collet to around 70mm above the table. Figure the mill will stick out 10mm and this left a maximum working diameter of around 32mm. Realistically though maybe 25mm. Even that would be hard to do because when you try to work that close to the limits you inevitable end up hitting the limit switch by accident when the machine moves to clear the workpiece. So I needed to extend the z axis with more movement. I had actually bought a taller z axis already but had to get bearings for it to ride on the 12mm rods.
I ordered the GRBL controller for the 4th axis and the 4th axis itself, and the bearings for the 12mm rods. Once I got all the parts I switched the Z axis to the longer one. This went from 50mm to 85mm of travel. Here's the old on the machine vs the new Z axis.
Then I swapped out the controller board and hooked up the 4th axis. At this point I ran a test just to see if the 4th axis worked. It took a few tries to figure out the software but it did work. Here’s pic and short video of the test. This is a piece of 1 14" Cherry dowel.
Continued on next post.
I’d like to preface this with a little note about me. I have an electronics background, a machining background, and I have been using AutoCAD professionally for years. I definitely don’t recommend doing this but it’s the journey I went on. I did it as much just for the gratification of pulling it off, as because I needed it. So I wouldn’t recommend doing this but I’ve enjoyed it. One of my main goals was to be able to produce a windless in an easy fashion. So how do you make a $800 windless? well follow along.
I'm not going into software discussions in this thread. I will say that I use Actcad which is a inexpensive version of Autocad, Fusion 360, Mudbox, Meshlab, and Deskproto. This machine uses a 4 axis GRBL controller. It's been discussed in several other threads on this site for both milling and 3D printers. The machine is the easy part. Learning and using the software is the big challenge. At any rate that isn't what this thread is for so I'm not going into any more of that.
A few years back I bought a run of the mill 3018 CNC machine because I wanted to try it out. I bought a Fox
3018Pro-SE. Like this.
At the time I was messing around with radio controlled airplanes and saw where a guy cut parts from carbon fiber sheet to create a wing folding mechanism for an F14. Long story, but bottom line is he was successful so I thought I would give it a try. He had talked about the power of the spindle so I bought an upgraded 300 Watt spindle at the same time. That was a good decision.
When I got it is was a pretty decent looking package but the entire z axis assembly that holds the spindle motor was a piece of printed plastic. The supports on the sides were some type of acetal material. The 10mm diameter rods were very flexible. Basically to make a long story short it was really easy to flex the tip where the tool mounted. Easily a 1/4” of play so there was no way this was going to do much in the way of accuracy at any reasonable cutting rate.
I had a 7x14 mini lathe and a manual benchtop mini-mill (variant of the Sieg X2) for quite a few years. I successfully converted the mill to CNC to make shock towers and chassis for RC Cars. I learned a lot during that project and the years of machining in the garage. One of the biggest things I learned is the practicality of using a mill to cut a particular material has a lot to do with the rigidity of the machine. The sieg would cut stainless steel pretty well but you had to go very slow. Once again to not go into too much detail, but if the machine starts to flex, the tools chatter and or break, the cuts aren’t accurate, the motors will lose steps, etc.. This forces you to slow down. There is a point at which you have to go so slow that it becomes impractical. I saw a beautiful relief carving done on 1 of these 3018 machines. It was very impressive but the guy that did it said it took 12 hours to machine, OUCH!.
At this point I knew this machine wasn’t going to do what I was hoping for. Not being one to give up particularly easily. I decided to scrap the basic frame and z axis of the fox alien. I basically stripped it for screws and miscellaneous hardware. I looked around and at that time they had just come out with the 3018 Pro Max. It is all aluminum, has 12mm rods for the bearings, 10mm aluminum plates for the sides and front and back and twice the size of extrusions for the rest of the frame. It also has more powerful stepper motors. They still sell this machine on Amazon for $277 USD. It's reasonably robust and is truly capable of relief carving at an acceptable level. Here’s a photo.
I used my 3D printer to make mounts for the limit switches that came with the Fox
and put limit switched on the Pro Max. Limit switches are something you can get by without but you will repeatedly ram the machine into the stops and this really isn’t good for the motors and electronics. I’ve somehow managed to never damage a motor but I’ve sure strained the heck out of them. Throw in the fact that they can be used to automatically home the machine and personally I wouldn’t run a machine without them. So here is how the 3018 looked at that point.
So time went by and I never did anything. Summer happened, and I wasn’t flying planes due to camping trips etc. This machine sat up on top of a cabinet in the garage for about a year and a half and I never really did anything with it. Last fall I was asked to help some friends with a model ship project. I actually spent a lot of time thinking about it before I finally decided to go ahead and do it. After I decided I would do it, I was asked to do something I’ve never done, and that was create 3d sculptures for the ship.
I definitely had my doubts about pulling that off, but figured I might as well try. After a couple weeks of trying I was literally preparing to send an e-mail explaining that it is way beyond my abilities. When I had thought about a different approach. I had a breakthrough and figured out a way someone with crazy tech skills; but weak artistic skills could do sculpting. That is a whole other very long story. The bottom line is. I did eventually manage to figure out a way to do what I consider passable relief carvings at least.
That got me to wondering If I could reproduce them on the 3018. My first try was with a 1/16” ball end mill and here was the result. Not great detail but it actually worked. A 1/16" endmill at this scale is a roughing tool, way to big to do much detail.
Then I did a little more work on toolpath and speeds and feeds and produced this carving. I showed these in Shota’s 3018 thread. The video is the top piece being cut in hard maple and the lower is Alaskan Yellow Cedar. These are finished with a 1/32" tapered ball end mill.
I was encouraged by this test to the point I really wanted to try and set this up with a rotary axis. I looked online for a controller that would work for a 4th axis and it turns out they do make one, so I hunted around for a 4th axis that would fit the machine. The smallest 4th axis I could find had the center of the 4th axis 44mm above the table. The Pro max had 50mm of travel. So I did some measurements and if the spindle was all the way up in the mount. It could lift the collet to around 70mm above the table. Figure the mill will stick out 10mm and this left a maximum working diameter of around 32mm. Realistically though maybe 25mm. Even that would be hard to do because when you try to work that close to the limits you inevitable end up hitting the limit switch by accident when the machine moves to clear the workpiece. So I needed to extend the z axis with more movement. I had actually bought a taller z axis already but had to get bearings for it to ride on the 12mm rods.
I ordered the GRBL controller for the 4th axis and the 4th axis itself, and the bearings for the 12mm rods. Once I got all the parts I switched the Z axis to the longer one. This went from 50mm to 85mm of travel. Here's the old on the machine vs the new Z axis.
Then I swapped out the controller board and hooked up the 4th axis. At this point I ran a test just to see if the 4th axis worked. It took a few tries to figure out the software but it did work. Here’s pic and short video of the test. This is a piece of 1 14" Cherry dowel.
Continued on next post.
Last edited:
