Cheap and easy linear supported rail

Some of the very largest – and coolest – CNC machines use supported linear rail for their movement axes. For any home tinkerer trying to reproduce these supported rails, the problem of cost comes up very quick; these rails can run over $100 for just a few feet. [Michael] came up with a great way to build his own supported rail so he can build his very large CNC router.

There aren’t many tools needed to build [Michael]’s rail. He put a 90° notch in a 2×4 to support his 25mm rail, and clamped it down with a piece of plywood. After drilling a 5/16″ hole every 12 cm, he tapped these holes out to receive 3/8″ threaded rod. Yes, we also hate the mix of metric and imperial units in that description, but the results speak for themselves.

The now-supported rail was mounted to a piece of MDF with a few bolts and washers. MDF isn’t the most dimensionally stable material, so [Michael] will be covering the whole thing in a coat of epoxy very soon. Now, he’s one step closer to his gigantic CNC gantry router.

You can check out [Michael]’s demo video after the break.

[youtube=http://www.youtube.com/watch?v=bGbWBDtzydw&w=470]

Comments

  1. Simonious says:

    Very nice, I can’t help but think that after all that effort you might have come out ahead buying the supported rails, but still, very nice results.

  2. Ant says:

    Hmm, it makes quite a ruckus and does seem to bind a bit. What kind of bearings are used?

    • Tony says:

      The bearing are the standard ones you use for supported rail, $5-$10 each.

    • Michael says:

      25mm Recirculating ball linear bearings & yes the noise is normal. They are lubricated with oil and are about 13$ each on ebay. No they are not binding, friction is from the 8 rubber shaft wipers, 2 per bearing.

  3. John says:

    I had to stop watching the video just before the end. I saw it go back, then forth… then back, forth, back, forth… back, forth… back… forth, and then back. I’d hate to never know what was in the last 5 seconds of footage. Can someone fill me in? 😀

    Nice build by the way! 🙂

    • Vonskippy says:

      Yes, it went left, then it when right, then it went left some more, then it went right, then a bit more right, then all the way right. Then… it went left again… (well I won’t wreck it for those who haven’t had a chance to watch).

      I could have watched 24 minutes of that.

      It has Sitcom potential written all over it.

      //kudos for not including some hideous death metal or cRAP music background noise//

  4. valdas says:

    Same thing i did only with smaller rail.

  5. stevetronics says:

    Gah! Those bearings sound like death! What happened to them? Are they lubricated?

    • Michael says:

      This is the normal sound of these bearings…
      It will not be heard over the cutting tool when the machine is running…

      • suitexperience says:

        I apply the same theory to my car. I cant hear the hideous noise from my near death wheel bearings if i turn up the radio 😀

  6. bootstrap says:

    How are the rails made parallel? That is the big problem with linear motion, everything must absolutely square or it binds

    • Michael says:

      The first rail was bolted down and then adjusted for straightness with the straight edge of a level.

      Then the Z carriage was added and the second rail was aligned by the Z carriage itself. Each bolt of the second rail was tightened as the Z carriage was slid from one end to the other.

  7. Mikey says:

    “No special tools needed, just a straight edge” … and then he goes on to show pictures of his drill press, tap & die, etc…

    Meanwhile, how is this even cheaper? You still have to buy the rods and the bearings which are the most expensive parts… not to mention the fucking solid steel custom cut gantry shown riding on the rails…

    So, I guess the frugality of buying expensive parts… and then just using them… is escaping me here.

    • Michael says:

      If your going to quote me, please get it right…
      “required no precision equipment other than a straight edge”
      A drill press is not precision nor special, they cost less than $200.

      It’s about 3 times cheaper than supported rail.
      The bearings are about $13 each. The rail was around $90 total.

      The Z carriage is not steel, it’s aluminum. Steel would have been cheaper and would work just as well. The holes to locate the bearing blocks are not precision drilled, there is a couple of mm of movement when it’s not bolted down.

      And mind your language, this is a public forum.

  8. Fallen says:

    That sounds awful. Is that normal for linear bearings?

  9. John says:

    Speaking as someone who has worked on linear motion devices before, I’m willing to bet that only the softness of the MDF is letting the thing move at all. The real nightmare is trying to get the rails perfectly parallel – unless the carriage has so much slop that you shouldn’t bother with rails anyway.

    • Michael says:

      I don’t have a dial indicator to check parallel but I will borrow one and make a follow up video.

      Every bolt is very tight and it’s not binding. And there was zero noticeable slop in a single bearing on the rail so I’m guess the method I used worked to bring the two rails in parallel without the use of precision equipment…

      A dial indicator will be the best test. I’ll post a link when I get a chance to test it.

  10. Miles says:

    Where did you get the material for the rail? That tubing looks fairly specialized.

    • Michael says:

      From the local metal shop in my small town, it’s pretty common stuff, almost every metal shop will have it.

      It’s solid not tube, and it’s called precision round rail. IIRC, it was $278 for 24 feet, or about $4.50 per pound.

      The section in the video used about $90 worth. The remainder will be used on the X-axis and the Z-axis.

  11. Alex says:

    How is $120 “cheap”?

  12. Chris says:

    I’m a software/electronics guy, with a bit of woodworking experience. Mechanical stuff and metalworking generally intimidate the heck out of me.

    So clear explanations with photos that don’t assume too much experience on the part of the reader, like this one, are gold to me. I’m pretty sure I could successfully make a supported rail after viewing this; but even if I don’t, some of the techniques shown could be used for other things. I’m definitely saving this one for future reference.

    Thanks Michael!

    One question. I assume the holes in the MDF/channel must be drilled a bit larger than the actual diameter of the threaded rod, to accommodate small inaccuracies. What hole diameter did you use?

  13. hojo says:

    doesn’t seem remotely comparable to an actual supported linear rail, and probably not much cheaper either. For the difference, I vastly prefer the precisely machined support of a known and consistent height.

    • Michael says:

      When I priced linear rail for the lengths required, I found that supported round rail was about 3 times more expensive than unsupported rail. Square rail was even more. As well, the supported rail was not available locally which would have incurred large shipping costs considering the lengths required (Y is 4′, X is 8′).

      I then considered buying the rail support separately, cutting it into short sections and spacing it along the rail. This would have been a bit cheaper, but it would require a precision surface under the precision support.

      With the 3/8″ thread rod supports, I can adjust the height individually to offset variations in the MDF.

  14. Ian says:

    the main challenge I see here (having worked on CNC mills from a maintenance perspective) is flex…
    you appear to be building a large gantry for cnc milling, but combined side-to-side flex due to using wood instead of steel is going to introduce positioning errors. they may not be large (and depending on your tolerances may not be noticeable at first) but once you start driving the table around with the weight of the stock you’re going to mill… then there is the issue of the milling itself… if you’re going to mill light easy to cut materials like plastics etc you may be ok… but as soon as you start trying to cut steel your milled surface is going to move away from the cutter perpendicular to the travel on those rails.
    if you’re attempting to machine parallel to the rails the milled surface is not likely to be smooth. (lots of chatter)
    however if you’re just planning on fly-cutting the upper surface of a material you may be ok as you will overlap the passes.. (I’d still make sure your direction of travel is inline with those rails and not perpendicular though)
    not trying to “rain on your parade” just hoping you are compensating for all these possibilities in your design…

    • Michael says:

      Hi Ian,

      The goal for this machine is to cut wood and at a slower pace, aluminum, machining steel is well out of the design scope. Comparatively, I have an CNC Sieg/Syil SX3 dovetail mill and rarely/almost never use it for steel.

      I agree, MDF by itself is not rigid at all. The MDF forms a torsion box with a pair of 1.5″ thick walled steel channels. You can see the complete gantry design here: http://n0m1.com/wp-content/uploads/2012/03/CNC-gantry.jpg
      You can also see a bit of the steel channel and torsion box structure in the video.

      The real benefit of using MDF, besides being cheap ($20 for a 4×8 sheet of 3/4″), is vibration damping. Alternatively, I was considering steel reinforced cast epoxy/granite but for this size it’s rather expensive and not very forgiving to mistakes.

      At some point in the near future, when I get my hands on a dial indicator, I’ll set the Y-axis up like a bridge (supported on the ends), load the tool plate and calculate the actual deflection. This will allow me to figure out practical feeds and speeds for a given material based on tool deflection.

      Michael

  15. The commetns are more wonderful.

Speak Your Mind

*

Related Hacks in cnc hacks

  • Mapping the surface of a video game moon
  • Pleasing results from a CNC mill project
  • And so the deluge of resin-based 3D printers begins
  • Hackaday Links May 9th 2012
  • Chocolate extrusion printer is halfway to making s'mores