Mini CNC Mill Goes Horizontal To Reuse CD Drives

Here at Hackaday, we pride ourselves on bringing you the freshest of hacks, preferably as soon as we find out about them. Thanks to the sheer volume of cool hacks out there, though, we do miss one occasionally, like this e-waste horizontal CNC mill that we just found out about.

Aptly called the “CDCNC” thanks to its reliance on cast-off CD drive mechanisms for its running gear, [Paul McClay]’s creation is a great case study on what you can do without buying almost any new parts. It’s also an object lesson in not getting caught in standard design paradigms. Where most CNC mills mount the spindle vertically, [Paul] tilted the whole thing 90 degrees so the spindle lies on its side. Moving it back and forth on a pair of CD drive mechanisms is far easier than fighting gravity for control, and as a bonus the X- and Y-axes have minimal loading too. The video below shows the mill in action, and it’s easy to see how the horizontal arrangement really helps make this junk bin build into something special.

We think [Paul] did a great job of thinking around the problem with this build, and we’re glad he took the time to tip us off. Apparently it was the upcoming CNC on the Desktop Hack Chat that moved him to let us know about this build. Here’s hoping he drops by for the chat and shares his experience with us.

 

22 thoughts on “Mini CNC Mill Goes Horizontal To Reuse CD Drives

  1. Horizontal spindle and mounting bed makes it tough to load very heavy work pieces, but if you’re not going to cut huge stuff (and on this mill that’s clearly the case) a horizontal spindle has a huge pile of advantages for helping clear chips. Slant-bed lathes manage the same thing, with less need for chip removal than mills, and they’re nearly ubiquitous in CNC lathe design now.

  2. CD-drive CNC machines were cool 15 years ago, but now that you can just get a bit of aluminum extrusion and make an actual tool instead of a toy, they seem a bit outdated.

      1. Agreed. I have two end of life computers sitting waiting to head to recycling. I also have an old disused dremel that works intermitently. I would love to give this one a try and see if I can pull off even remotely similar performance to this excellent hack.

  3. Minor question. I assume this is behaving a bit like a 3d printer in reverse. Cut, step, cut, step, cut, step. Why no fully 3d finishing when it seems like the tool should be able to do that? It was “z-hopping” (y-hop??) to move between pieces in the cut. A fully formed 3d path should be pretty trivial to finish as expected.

    1. He learned that unless the tool can take a ‘bite’ and make a clean chip it will heat (and melt; in this case) the work, so he had to reprogram so that the tool could bite off what it could chew so to speak.

      The cd drive motors had trouble making enough force to get the tool in. The article is quite well written, worth a look.

    2. Hi Hal,

      Sharp question. Answers include (1 of 2): I didn’t try.

      I rationalized not trying this way: Moving the Z axis requires nearly all the torque the tiny stepper can produce. In that condition, the mechanical angle of the output lags the “magnetic angle” (for lack of a proper term. ?.) by most of a step (iirc). When reversing direction the lead screw doesn’t move at all for most of two steps while the “magnetic angle” unwinds its first lead and gets ahead of the screw in the new direction. One step = the 0.15mm layer step-down used in the example pieces. So that’s a lower bound to hysteresis even with zero mechanical backlash between the lead screw and tool tip. Then getting the mechanical lash down while shifting the heavy load would be its own challenge which I passed up given the limit to what that would be worth even it perfect.

      Working with greater Z range in thicker material would reduce the impact of all that. Another project mentioned in the writeup does that with Z-varying tool paths in something like floral foam. I fixated on the possibility of cutting harder materials, so patience limited depth of Z range.

      Looks like I’ve just busted myself down to not trying due to impatience. Sigh.

      Anyhow… answer 2 of 2:
      Because doing that
      here: https://hackaday.io/project/174370-cheap-small-cnc-mill-formula-1551-for-now/log/182669-parallel-finishing
      and here: https://hackaday.io/project/174370-cheap-small-cnc-mill-formula-1551-for-now/log/182748-backlash-chatter-and-fine-parallel-finishing
      instead!

      1. Paul thanks for the explanation. I hadn’t considered the backlash issue and the precision that would require.

        And nifty design on the other one. I didn’t realize you had multiple designs of the mini mill. For simple quick parts this one should be a functional tool. And a great reuse of the tech. Thanks for answering my question so fully.

  4. Author: “This project demonstrates the value of ignorance. If I’d known half of what I’ve learned about why this is impossible, I wouldn’t have considered trying.”

    I don’t understand this statement. He says, “…why this is impossible.”, but then shows it is possible. Maybe replace “impossible” with “a bad idea”?

    1. I think it refers to CNC forums where unless you have a granite epoxy or full aluminium CNC, it is impossible to do any good work.
      I’m exaggerating a little bit, but you get the idea!
      Cheers

    1. With some interpretation: Yes!

      For this horizontal arrangement, taping a piece of paper to each side of the base (filling the right angle between the horizontal and vertical boards) suffices to catch practically all the chips. And less energetic chips simply fall off the workpiece.

      I’ve lately come to appreciate what a free win that was after trying a vertical configuration where chips pile up on top of the workpiece without limit: https://hackaday.io/project/174370/log/182589

      When i get tired of sitting there with a vacuum hose in hand I’ll have to figure how to attach it to the frame (the tool moves but the point that needs vacuumed does not).

  5. Osgeld may lack tact, but isn’t the only one with that sentiment. The Web would be a lot faster and simpler to use with a few tenets learned in the ’90s (and before!) and that everyone has unlearned in this Web-2.0+ era, where HTML is replaced with obfuscated Javascript, and designers try to squeeze in every fad from distracting mouse-over zooms and animations to image crops and vignettes and lazy-loading flicker, none of which can be disabled with browser settings without making most of those “websites” unusable.

    Not to mention the side-channel implications of running all these random programs on your computer. Has nobody learned from Spectre?

    Hackaday.com itself (not .io) is one of the few islands of usable Web left, with simpler styling and actual HTML, and sometimes the project pages it has links to are even simpler. I dread the day this site’s content gets submersed in the flood of flashy fads.

Leave a Reply to smellsofbikes Cancel 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.