This seems like a noob question buried under a nut-case idea, and I am sorry if it is. How does the part-time maker build compact (maybe even portable) assemblies that can be used for multiple processes? Could a laser welder also be a 5-axis mill in its spare time? Could an interchangeable head turn it into a plasma cutter? Can these home-built machines operate at industrial accuracy? I kept coming up against weight and drive-slew problems based on inertia models and hardware that are highly-calibrated and proprietary. I keep coming back to a long-term repeatability problem with drive systems and frames supported by off-the-shelf parts. Then I stumbled across a video of a tramming tool making a G-code model of a small cast part, and it seemed to be teaching itself tool-path and slice as it mapped dimensions of the part. I don't think we have to choose between accurate and affordable (speed and cost seemed tied together). I think suitable parts to make CNC structures be highly accurate, reasonably priced, and with almost no human adjustment during service life are right in front of us already. I just haven't looked in the right part of the problem. Somewhere in this haystack I believe there's a simple needle we can all use to stitch easy, affordable, and accurate closer together. I've been evaluating kits for 3 days solid. So far, nothing. Thanks for taking a minute to read a madman's rant.
The answer is yes. A multi-purpose machine can be made but you have to consider limitations and compromises. I have high accuracy (enabling rotary engraving of text less than 0.5mm height) and high torque (enabling V-Carving of hardwoods 12mm deep in one pass). My compromise is speed in so far as my maximum, reliable, feed-rate is 4500mm/min and as the same machine is also used for laser work having double this speed would be nice. Changing my ball-screws from 5mm pitch to 10mm pitch would give me the 9000mm/min feed-rate but at the expense of accuracy. Using the existing ball-screws and changing my stepper motors for servo's would also give me the 9000mm/min but at expense of cost. As you say there are always three choices but we are only allowed to pick any two. Tweakie.
Tweaky, what brand of ball screws did you use? I think of using them too, for a belt system is not as reliable (because of the stretching of the belt). Found this a couple of days ago: http://www.thomsonlinear.com/websit...and_lead_screws/ball_screws/metric_series.php But I think it will cost pretty much... (In the menu on the left, there is also a tab for the Imperial sizes) Do you have a cheaper solution?
Unfortunately, I have no cheap solution. Like everything really but there are ball-screws and there are ball-screws with some of the cheap Chinese made products being extremely dire indeed. My ball-screws were made by THK in Japan, unfortunately quite pricey, but the amazing thing is that the associated angular contact thrust bearing blocks cost almost as much as the screws. Tweakie.
