Cobbling Together A CNC Mill

[Howard Matthews] never throws anything away, and because of it he was able to build this CNC mill using parts that he already had on hand. He pulled stepper motors out of broken stage light, precision rod and bearings from old dot matrix printers, and other various bits from his junk bin. We’ve seen [Howie’s] handi-work before, and this project is just as fun as his Land Rover’s replacement speedometer. Some highlights include manufacturing the nuts for the precision rod, and building a rail system for the bed of the machine. The latter looks a bit suspect, as any milling debris on the rails will cause you Z-axis problems, but now that he has bootstrapped a working mill perhaps he’ll machine an upgrade.

Update: Fixed the link, added video after the break.


22 thoughts on “Cobbling Together A CNC Mill

  1. Can’t you just use central heating pipes for guiding? Although then you’d need to find a suitable bearing, but that’s not impossible.

    @brian yes, and I see that on more sites too, but not all of them oddly enough.

  2. I really don’t mean to sound like an ass, but compared to my Taig, this thing is flimsy. The body being made of wood, and with the torque of only a dremel tool, I’d only machine plastic and wax on this.

    Don’t get me wrong, please- this is quite awesome, given what he had laying around. It will be lucky to machine anything other than wax beyond 0.005″, if even that, but it would work just fine for wax & plastic.

    If you want to machine metal, you can’t have flex in the body. You need rigidity, or you will have lots of deflection with heavier things like aluminum even. Getting rid of deflection is how you machine past 0.005″.

    If he even screwed together cross braces behind the boards, and sandwiched more boards, he might be able to get around 0.001″ in wax.

    This is inspiring, as really, unless you are working with very precise stuff, this kind of machine would allow all kinds of closely made stuff in wood, plastic, and wax for the homebrew people. I bet it would make great circuit board cutting too.

  3. @drew: yes you sound like what you feared. An imperial ass at that for using medieval units too. Perhaps you mill a lot in steel, but not everyone measures with your yardstick. Are you complaining too you can’t cut steel with a wood-saw?

  4. @Drew although your balanced comment is valid, search for home built CNC and you will be amazed to find people happily cutting there way through Aluminium will very small tolerances. Cutting pockets etc. Very cool i want to build one.

  5. To all the “it’s flimsy!” people out there. I would LOVE a machine like this.

    I have no need to cut metals, or even heavy duty wood. My main target would be balsa wood. And maybe PCBs.

    Just because it doesn’t suit your goals doesn’t make it a bad design.

  6. @Drew: Yeah, it’s strange looking at hobbyist CNC machines coming from the real thing. A lot of them are only designed to mill machinable waxes and polymers. After looking at the guerrilla CNC guide, it seems that it’s common for these guys to use the mill to make molds for resin pouring, which means they’ll never have to mill metal at all.

    I would have no use for such a thing, but for their specific uses, they are quite effective, and the cost of entry is very low.

  7. @drew

    Stiffness is an important consideration, however it isn’t the deciding criterion for the accuracy of the mill, in fact to equate the accuracy of the machine to its stiffness is to ignore 90% of what makes a machine accurate, and to equate the precision of the machine to the number of sandwiched boards and cross braces used is on the brink of nonsense.

    The accuracy of the cut depends on the feedrate (as does the deflection of the tool), the cutting tool, and most importantly calibration and quality of construction of the machine – he could make the entire thing from tungsten carbide and it still couldn’t cut with 0.005″ accuracy/repeatability if the leadscrews have uncontrolled backlash, the axis aren’t aligned, or the bearings have any play, amongst other things.

    With a well designed machine, you can can achieve high accuracy (beyond 0.005″) in aluminum with what you would consider insufficient stiffness – you just have to be patient. Since the forces applied to the machine are related to the feedrate you can cut slowly which results in much less force on your cantilevered tool, therefore less deflection and an increase in accuracy, at the penalty of parts taking forever to finish – which doesn’t bother many hobbyists. What bothers most hobbyists is when people find the need to compare the homebrew machine to their “Taig”.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.