Hello all, I'm de-lurking after reading both this thread and the hardware store cnc thread start to finish in the last 2 days. Let me start by saying THANK YOU to everyone who has posted here. It was exciting to read what you have done so far, and your hopes for where you were headed.
I was REALLY hoping that either of these 2 projects would have resulted in a "standard" design, based on best practices. I'm finally ready to start my own project, and was hoping for some distilled down, tested designs. That way I could get busy making chips fly, instead of re-inventing the wheel. I decided to post in this thread because this is more the type of machine I'd like to build. It's ok to limit oneself to hardware store items, but it seems kind of artificial, especially when so many better items are available from other stores and the web. (The disagreement about whether to allow square steel tube was the killer for me, it's readily available at most home centers, and every city in the US has somewhere you can buy steel. Worst case it's available mail order.)
There are so many great projects in the worklogs, and so many great ideas in this and the hardware cnc thread, but sometimes the choices are overwhelming. And some of the gaps are just mind-blowing. Like WHY to choose steppers over servos, WHAT actually goes between the CAD drawing, (the G-code) and the drivers, to actually make the machine DO things. Not to mention why choose different methods to drive the axis' like direct coupling, reducing gear boxes, threaded rods, rack and pinion, belts and pulleys...
I'm not suggesting this is the right place for a primer on how and why, but I would think it's the right place for some discussion about approach and choices when designing a "reference" system.
I have fairly sophisticated building ability, and access to a well equiped workshop, although no real machine tools like mills or lathes. I feel confident that I could execute any design presented here. And I can read the worklogs and borrow this good idea from here and that one from there, but the whole point of these threads seems to be to learn from others experiences and distill that knowledge down into what business calls "best practices." To that end, we would need to avoid what I think happened in the hardware cnc thread, people got bogged down in detail, and lost sight of the whole. Would it not be better (and reinvigorating to the thread) to identify the "big chunks" and work in parallel on them instead of working in tiny detail, each item in series?
For example, every machine needs these "chunks" or modules:
supporting structure/ frame
work-holding table (x axis?)
spindle/ router holding structure (y and z axis?)
Then look at choices- moving table or moving gantry? what support for the moving parts, shopbot style? rails? tubing? (square? round?) angle? Purchase or make?
What drive mechanics? threads, pulleys, rack? Linear induction motors?
What drive electronics? What are good choices for the guy who doesn't want to learn electronics and build his own possibly dangerous power supplies/ controllers.
What type of material to cut and what does that require in terms of hold downs, table structure, chip removal, cooling lube? How does that determine the structure of the other parts?
then look at the pieces:
support frame, made of what? how ridgid? how big?
work holding, made of what? move table or gantry?
etc.
Then apply the whole groups ingenuity to solve the problems out of commonly available materials, using simple techniques, with the goal being not the worlds sweetest homebuilt CNC router, able to hold .00001 of an inch over 10 ft of travel, perfect and innovative in every way, and oh yeah only costs beer money, but to provide a CNC that can hold real world tolerances, be built by a reasonably handy person, without esoteric materials, tools, or skills, with the goal of actually finishing the machine and cutting stuff!
So, is there any interest in covering some of this ground again, and trying to restart this project? It's possible there isn't really a need for this sort of "reference design." Maybe the people who are drawn to build their own machine tools are also the kind who need to do everything "their own way." But even then, I think there would be great value in having a starting point for basing your own design and custom details on. A design that is "good enough" for 90% of people, that uses proven techniques and materials.
I'd like to hear your thoughts on that possibility too.
With respect,
zuk