I had been quite excited when I found out about CNC machines and did a lot of research on them. After a while I started to look at the DIY reports about people making their own CNCs. While not the best out there it seems like making a CNC for wood carving out of wood was possible. I let the idea sit for quite a while but have finally bit the bullet and ordered parts for my first attempt.
The CNC has four main parts, the software, the frame, the electronics and the spindle. For software I used LinuxCNC. For my frame I used v bearings and steel L brackets supported by 2x4 lumber. This was driver by 1/2" threaded rod. For electronics I grabbed a 3 axis kit from ebay (link here). For the spindle I used my Proxxon rotary tool with 1/8" end mills.
The cost ended up being around $600, with half of that being the electronics package. The v bearings were also quite expensive but I knew I could use them on other things if the CNC did not work out. After a few weeks of cutting and putting things together I had a working CNC.
While far from perfect this CNC allowed me to do a few engraving projects that were small and worked well, as well as get used to using a CNC. So after a year of poking at it I decided it was worth it to start looking as building a new CNC.
Looking back there are a few things that I would have done differently, the first is get better tools. I was extremely limited in what building tools I had, (hand drill/jig saw/hack saw) so the frame and linear rails are very roughly built. Now that does prove you can create a CNC with basic tools but the quality will suffer. Second is making my gantry smaller. I had some dreams of grandeur thinking I could have a CNC and 3D printer in one. This is a bad idea and it will not work as CNCs are too slow. A CNC really needs to have a gantry height that will just fit what you are working on, the taller it is the less stiffness you will have, which was a problem in this one. Finally I would go with better linear rails. I liked the v bearing idea as it is easy to extend if you need longer rails, but it just is not accurate and stable enough the way I did it. I am thinking square tube embedded in a v channel in the wood would be a better approach, however I think i will go with SBR rail on my next design as I just do not think the tensioning on bearings is easy to work around.
The best thing about this whole project is that I was able to find good uses for a CNC and learn the software side as I used LinuxCNC with CAD software as well as some straight Gcode designing. if you think you may want a CNC I would highly suggest going for a simple DIY version first, looking back a simple drawer slide version would have been cheaper and just as accurate if not better, but as you can reuse the electronics and motors a simple CNC will get you started and from there you can figure out if it is worth investing in a better CNC after that.
The CNC has four main parts, the software, the frame, the electronics and the spindle. For software I used LinuxCNC. For my frame I used v bearings and steel L brackets supported by 2x4 lumber. This was driver by 1/2" threaded rod. For electronics I grabbed a 3 axis kit from ebay (link here). For the spindle I used my Proxxon rotary tool with 1/8" end mills.
The cost ended up being around $600, with half of that being the electronics package. The v bearings were also quite expensive but I knew I could use them on other things if the CNC did not work out. After a few weeks of cutting and putting things together I had a working CNC.
Now I knew starting out that I would either improve on this CNC or build a new one after getting some experience so I was not expecting the greatest accuracy or performance. And my expectations were right. While the cnc was ok in accuracy my linear rail setup was not the best. Part of this stemmed from only having a few basic power tools and not being able to grind the rails smooth. So the cuts on it tended to wiggle a bit as the gantry moved. My choice of threaded rod gave it a lot of moving power, but the trade off was the speed was quite slow, close to 5 in/min which is quite slow. I had no trouble with a small end mill but a full size router would likely have burning issues.
Looking back there are a few things that I would have done differently, the first is get better tools. I was extremely limited in what building tools I had, (hand drill/jig saw/hack saw) so the frame and linear rails are very roughly built. Now that does prove you can create a CNC with basic tools but the quality will suffer. Second is making my gantry smaller. I had some dreams of grandeur thinking I could have a CNC and 3D printer in one. This is a bad idea and it will not work as CNCs are too slow. A CNC really needs to have a gantry height that will just fit what you are working on, the taller it is the less stiffness you will have, which was a problem in this one. Finally I would go with better linear rails. I liked the v bearing idea as it is easy to extend if you need longer rails, but it just is not accurate and stable enough the way I did it. I am thinking square tube embedded in a v channel in the wood would be a better approach, however I think i will go with SBR rail on my next design as I just do not think the tensioning on bearings is easy to work around.
The best thing about this whole project is that I was able to find good uses for a CNC and learn the software side as I used LinuxCNC with CAD software as well as some straight Gcode designing. if you think you may want a CNC I would highly suggest going for a simple DIY version first, looking back a simple drawer slide version would have been cheaper and just as accurate if not better, but as you can reuse the electronics and motors a simple CNC will get you started and from there you can figure out if it is worth investing in a better CNC after that.
