High-Precision Air Bearing CNC Lathe And Grinder

You know you’re in for a treat when you are told that a lathe which can reach a resolution of one micrometer (1×10−6, a millionth of a meter, or 0.00004″ for people who love zeros) is ‘not hard to build’.  This is one of the opening statements in this video by [Dan Gelbart], as he walks the viewer through the details of a custom CNC lathe which he built. (Video embedded below.)

As it’s a combined CNC lathe and grinder, it uses custom software he had developed specifically for the machine. Much of the high precision of the machine is courtesy of air bearings. All but two of the air bearings were made by [Dan], with the two surplus air bearings he used coming from machines used in the semiconductor industry.

The bed of the machine is formed out of off-the-shelf reference granite, to which the other parts are epoxied, providing a stable base with well-defined dimensions. Though perhaps a few light years beyond most DIY lathe efforts, [Dan]’s videos nevertheless provide a treasure trove of tips and information for lathe builders and users alike. Certainly worth a look.

Thanks [Drew] for the tip in the comments.

27 thoughts on “High-Precision Air Bearing CNC Lathe And Grinder

  1. The most singular feature of his videos is the selection of tools he uses. He makes a lot of use of things most of use don’t have, but are not particularly expensive, like a brake for sheet metal and a spot welder/stud welder. I don’t know anyone who can afford a personal water jet, but using materials that can be cut by a laser can do a lot. Keep an eye out for a small shear? Are there alternatives that work as well?

    Grinding in machining is something everyone should learn. Tool post grinders I have used are so well balanced and bearings so good (and motors) that they are nearly silent.

  2. great work, however working with a lot of higher end medical and aerospace production companies and seeing the hardware they have he is quite misinformed on some things. and I do not even know where to start but 2 big things come to my mind. First off EDM lathe’s are more precise and can handle a much higher tolerance than his is able to. Also the big thing they run into with machines that are that accurate is temperature control. Coolant, room temp, stock temp all must be kept extremely precise at all times during the machining process. even a degree difference f or c in these instances causes huge variances in production.

    anyway still great work

    1. You are correct about the temperature issue. True toolrooms were and are always temperature controlled, to the international standard of 68 degrees F. This is the temp that all gauges and measurements on all blueprints are actually supposed to be made at, as it was agreed upon.

      The EDMs I run use a special coolant to chill both the ballscrew motors and the part being machined to keep the measurements stable, and the die mills I run do the same with their ballscrew motors for the same reason.

      EDM lathe- please explain? I can quite guess how that would work, though I’ve never seen such a thing. I use my Agie SP 300 diesinker EDM to rotate at 50 rpm to burn ejector pin holes, and it orbits too…If you bend graphite, you can make it burn intentionally tapered holes spinning and orbiting… Are you talking about the electrochemical grinding tech Oberg uses?

      Glad people like Dan’s ridiculous lathe- id pay heavily for a usable book to make one (I actually have use for such a beast…)

    2. I run diesinking EDMs with rotation ability. I can imagine what you mean, but have never seen or heard of an EDM lathe. Can you elaborate?

      Oberg’s electrochemical grinders that are proprietary are the only thing that come to mind.

      Glad people like Dan’s insane lathe. I’d pay for a book that gave step by step construction plans on making one.

      1. I’m a machinist as well and have never heard of an EDM lathe either…maybe they’re talking about an EDM machine with a rotary axis? I haven’t seen anything like that myself and don’t know such a machine exists. They could possibly be talking about a machine with orbit capabilities for EDM thread milling too.

      2. The EDM lathe’s i’ve seen in the most general of setups are as follows. They have very high precision indicators both in front of and behind thecutting tool, and they are using a special carbon cutter(i think it is carbon) much in the way a sinker is used. They are very slow rotation but burn on the turned axis. I’ve always been informed or told they were edm lathes however they might actually be sinker edm turned sideways so they could do turning. I hope that description helps out and sorry no make/model numbers at all I just get to see the cool equipment they font let me play with that stuff :)

    3. “even a degree difference” – When I worked at Dan’s company (Creo) a few decades ago, I heard stories of such fine temperature control a person could not enter the room during machining. He knows high-end.

    4. @TheGamler: Obviously an EDM lathe doesn’t suffer from pushoff or part flexion.

      How do you bore complex ID’s with an EDM Lathe? – is internal thread cutting possible?
      (or would one need to revert to submerged EDM with a solid forming tool. -then tool erosion/deposition will vary accuracy.)

      True that “very high precision” isn’t possible without temperature control, but then the parts made will usually need to have tolerances which allow for fitment and use in varying atmospheric conditions. So beyond “a point” tolerances become academic, even matching parts will vary critical dimensions unless they have similar sectional thicknesses – ie. OD to matching ID.

  3. What temperature is the measurements at? At those tolerances unless an arbitrary higher temp is soaked in, your measurement is going to vary all over the place and definitely not repeatable

      1. why would 2 decimal places be an issue? I’m more offended when someone asks for 4 decimal places and they could have lived with 2. I was doing some grinding to a very silly thickness tolerance and another engineer designed a gauge to measure the thickness, but it was too big. That cost thousands of $ because it had tight tolerances. I replaced it with a bandsawn piece of scrap that worked great for $0. The other engineer told me I couldn’t let anyone see my gauge because it wasn’t respectable looking, which is fair.

        1. I get that too, and that’s even more aggravating. Engineers overtolerance things constantly.

          Why 2 place decimal upsets me? Because they usually ask me for very tight dimensions, but give me a print when I’m done that says it was only 2 place. They do the opposite too, give me a print before I start, but ask me to make a dimension tighter than the print.

          Basically, engineers being lazy and taking shortcuts, combined with careless and capricious application of tolerances is what actually upsets me.

        2. I agree! Engineers like to or don’t know better using tolerances. They learn a lot from machinists. Personally, I don’t like band saw cut finishes even thou it may be good enough. I’ll clean it up a little.

    1. Congratulations, you knew about it all of a few years ago before some others here. So did I and some others.

      That doesn’t make it any less incredible to lots of people who have never seen this before.

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.