Next Project – Yet Another Homebrew CNC Mill

CNC Mill Prototype

The idea of a machine that takes digital designs and creates real things has always been fascinating to me.  I like that I can go from an idea to an actual physical object with just a couple hours of work (or minutes in some cases), a little scrap material, and one of these machines at my fingertips.  I recently came across a laser cutter for sale in my area, and I had to get it for this reason.  I absolutely love what it allows me to do with my ideas, and I really see it as a "gateway" device for me.  The laser cutter allowed me to build a 3D printer from open source designs and make modifications.  Now I turn to my laser cutter to help me make the device that introduced me to these dream machines, the CNC router.

I've wanted one since high school and I've played around with the idea of making my own for some time now.  I decided that now is as a good a time as ever to try and create my own version.  Here's some of the goals I set up for this build and the reasoning behind them:

Reproducibility:
I've made my share of projects from found components, both bargains and the kind pulled from dumpsters, and while this makes each project unique, it often doesn't allow others to follow in your footsteps easily.  It also means that when the project breaks it often can't be repaired easily, if at all.  For these reasons, using primarily readily available parts is a must (you'll fine out why I use the qualifier "primarily" here in a second).

Cost:
I could go all out and get high quality linear bearings, ball-screws, anti-backlash nuts, and use milled aluminum components for construction, but why?  If I was going to do that I'd just go out and buy a good CNC mill made by people that know a lot more about mechanical design than me or buy a good mill and add steppers.   There are even a number of low cost but decent quality hobby CNC setups available, most notably the FireBall V90 from Probotics and the micRo from Lumenlab.  These units cost $600 for the mechanical components and $900 for mechanics plus steppers, respectively.  To make my design worth while, I've got to build a better system for about the same price, which its unlikely, or I've got to make a reasonably performing system that's significantly cheaper.  Half the cost of a readily available kit seems like a good goal to me.  This would mean that all the mechanical components of the machine should cost under $300.  I envision the final version being constructed from 3/4" HDPE, which I estimate will cost about $50 for a single machine.  If that is the case, all the mechanical bits should cost under $250 to meet this design goal.

Performance:
At the the very least the machine should be able to mill plastic reasonably well and preform other lighter load tasks.  Lighter load tasks would include using it as a simple draw bot, 3D printing using a plastic extrusion head, and PCB milling if possible.

Modularity:
This entire project will be a learning experience for me, and I want to be able to use the mill as an experimental platform of sorts so the ability to swap out components is important.  I'd also like to design a few variations of different subsystems that people could swap out depending on their desired use case.  For example, I plan on using Acme rod and hex nuts under tension to help with backlash, but to cut costs you could alternatively use 1/4"-20 all thread and a coupling nut and compensate for backlash in software.  Ideally you'd only have to swap out only a few parts to go from one setup to the other.

Bonus Points:
Just for fun, I'd like to design the mill such that all the machined components can easily be made on the mill itself once it's completed, much like the concept behind the RepRap.  Since I don't have access to a CNC mill myself, this means that I'll have to create a "bootstraped" version with the tools I have and then mill the parts for the final version on the bootstraped machine.

I've already started making some components and I'll hopefully have some stuff to share soon.

Share

Tags: