584,798 active members*
4,244 visitors online*
Register for free
Login
Results 1 to 16 of 16
  1. #1
    Join Date
    Apr 2006
    Posts
    23

    Homemade Linear Slide System

    Building my own CNC machine, for both plasma cutting and for slinging a dremel around, and after pricing ebay/used/new THK slides and competitive products, I decided to at least try making them myself.

    Turned out not to be anywhere near as difficult as I thought it would be. I hope I posted this in the right place; After looking at several forum names I guess it could fit in multiple places, but I chose this one "just because"

    Some bearings, some square stock, a few drill bits, and a digital caliper makes this easy work. Figured I'd share this since I see a bunch of threads about how to make such things yourself.

    Step 1: Making the saddle chassis (need four of these):


    Step 2: Making sure all the holes line up well enough to bolt together:


    Step 3: Machine two spacers per bearing (eight bearings total per saddle). Whit the bearings I selected the two spacers need to be 0.152-0.153" for a tight fit:


    Step 4: Bolt it together and hope it fits:



    Step 5: Make more!

  2. #2
    Join Date
    Jul 2003
    Posts
    276
    Looks nice but you can't support the rail with that design.. unless you leave out the bottom 2 bolts (think upside down U instead of an O)

  3. #3
    Join Date
    Oct 2005
    Posts
    612
    I like the way you have positioned the bearings around the shaft - much better than the popular angle design as you will not need to load the shaft to maintain bearing contact. You also have flat faces to mount the next component.

    Just when I thought it had all been done before another great idea pops up.

    Well done and let us know how they perform.
    cheers,
    Rod

    Perth, Western Australia

  4. #4
    Join Date
    Apr 2006
    Posts
    23
    Randy - You're are absolutely correct - the rail can only be supported on the ends because the slides (saddles) are fully enclosed. I'm doing this on purpose actually, because my original design was held together by gravity, and I wasn't happy with the outcome. The original design started off as a homemade pattern copier, actually:

    Old design still:


    Old design moving:
    http://frederic.midimonkey.com/cnc/IM001188.AVI

    Anyway, this machine as you can see isn't going to be that large, (30x21" work area, machine about 3" larger in both directions) so I'm hoping the rails won't flex because of the small size and reduced weight of the thing not being so large. If your concern turns out to be a problem, know that I can fix it with this design

    The design of my end supports for the rail and the leadscrews just below, has enough space in the design that I can install up to a 1.25" diameter rail, therefore I can gain more stiffness in the rails if necessary. The saddles as you can see in the pictures I posted already are adjustable - longer bolts and wider spacers - means I can ride the saddles on larger diameter material. The "rail" in the picture is a short section of 1/2" thin wall tubing I had laying around and used to illustrate how the saddles will mount, but the "real" rails are 1/2" diameter hardened round stock, which is by far stiffer than the thin wall tubing in the picture. How much stiffer? I don't know. I can bend the tubing against the back of my neck, whereas the solid, hardened round stock I cannot. I can barely flex it. But like I said I can increase the diameter of the rails up to 1.25" without having to machine new supports, make new leadscrews, and so on.

    While I do intend to make a dremel head eventually, the main purpose of this machine is to perform plasma cutting, which is an advantage - plasma cutting heads don't weight that much, and if operated correctly "floats" about 1/16" above the material being cut, so there is no contact that will cause resistance in the movement of the machine. So at least for plasma cutting, the rails/saddles only have to support it's own weight, plus the weight of the other axis' installed on top.

    One of the intermediate designs I came up with would have allowed for a supported rail, and still force the saddle to only move in the direction of it's assign axis - using three bearings per end, in a "Y" type of arrangement, looking down the length of the saddle. I made about six of these types of saddles before I finally accepted *I* was unable to make an accurate enough "Y" mounting arrangement so I went the easy route with a square, four bearing arrangement. Though I was making the "Y" style saddles by welding steel flats in a wooden jig I made, trying to get the 135 degree angles as accurate as possible. I just couldn't do it. If things do flex really bad, and upgrading the diameter of the rail for stiffness doesn't help enough, I may have to revisit the "Y" arrangement, and support the rail underneath with say, a 1/4" wide flat of appropriate height.

    If I had a milling machine at my disposal, making the "Y" arrangement would be easy - I'd just make them out of solid aluminum stock and machine the shape, and machine slots for the bearings, then drill through to tie it together.

    One of the things that was important to me, is to have the bearings supported by bolts that are in double sheer.

    For about a year I was looking for a used "all in one" lathe/mill combo, however being unemployed for the past 18 months didn't help me afford such a unit, so I ended up with a used Clausing lathe instead:


    Anyway, good eye Randy. Also, just in case it's not obvious, I'm not arguing, merely sharing my thinking at the time I started on this adventure. I realizing my writing style doesn't always facilitate that. One too many corporate memos I guess. Anyway, I very much appreciate the comments.


    Rod - having a flat surface to attach the other axis too is key, and I put a fair amount of thought into it. Randy's concerns above concern me as well, however as I stated above if there is too much flexing I have an upgrade path outlined to reduce or eliminate that.

  5. #5
    Join Date
    Apr 2006
    Posts
    23

    Progress as of 2006.05.25

    Drilling out the ends of the stainless, hardened rod to 27/64"


    Tapping to 7/16 course:


    Completed rail, with rail supports. I still have to chuck the end supports and skim them with a carbide cutter to make sure as both ends of the rail are the same height off the chassis. You can see the saddle floating in the middle.


    Completed second rail, with supports, still have to skim the bases of the four end supports. I'll do that in a few days once I swap lathe chucks.


    Anyway, back to making 3/16" ID, 0.153" long bushings

  6. #6
    Join Date
    Dec 2005
    Posts
    33
    Nice home made rails system.

    I may have missed it, but what is the diameter of the rod?
    And is the rod regular round? or the grinded shaft material?

    If one were to do something similar, what kind of bearing be appropriate?

  7. #7
    Join Date
    Apr 2006
    Posts
    23
    I'm using mild steel 1/2" diameter tubing at the moment (as in the above pictures), as a place holder while I'm still experimenting and engineering "on the fly". It will be replaced with hardened stainless solid 1/2" round.

    The small diameter seems okay so far, but it is a small machine, and the plasma cutter doesn't apply torque to the material the way a router would. It just hovers over the material (1/16" being ideal).

  8. #8
    Join Date
    Jan 2006
    Posts
    481
    Quote Originally Posted by midiguy732
    Building my own CNC machine, for both plasma cutting and for slinging a dremel around, and after pricing ebay/used/new THK slides and competitive products, I decided to at least try making them myself.

    Turned out not to be anywhere near as difficult as I thought it would be. I hope I posted this in the right place; After looking at several forum names I guess it could fit in multiple places, but I chose this one "just because"

    Some bearings, some square stock, a few drill bits, and a digital caliper makes this easy work. Figured I'd share this since I see a bunch of threads about how to make such things yourself.

    Step 1: Making the saddle chassis (need four of these):


    Step 2: Making sure all the holes line up well enough to bolt together:


    Step 3: Machine two spacers per bearing (eight bearings total per saddle). Whit the bearings I selected the two spacers need to be 0.152-0.153" for a tight fit:


    Step 4: Bolt it together and hope it fits:



    Step 5: Make more!

    hi midiguy732


    very nice concept with the linear bearing arrangments , keep up the good work

    cheers

  9. #9
    Join Date
    Apr 2006
    Posts
    23
    Thanks!

    Actually the machine is almost done. Got the pulleys yesterday, have to make the stepper mounts, and measure for belts. I'm behind on the pictures

    http://frederic.midimonkey.com/cnc-gantry2.html

  10. #10
    Join Date
    Jul 2006
    Posts
    3
    Very Nice!!! Thanks for the construction detail.

  11. #11
    Join Date
    Jun 2006
    Posts
    35
    I was thinking about as I would be to make a table x,y with this bearing/spacers slide system... hmmm ... maybe ?

  12. #12
    Join Date
    Jan 2006
    Posts
    481
    hi
    midiguy732 , i come across this skate_bearing rail setup some where of the net

    alot more maching required , but if someone is thinhking of making a larger plasma or router table needing more support for ridgity under the mail rails this could be a simple solution

    less bearings one side so only three bearings hug the mail rail allowing for a bridge arrangment so the skate_bearings( ballscrew ) doesnt hit the mail rail supporting feet .

    I'm sure even with ur setup midiguy732 , one could leave out one side of the bearings and fit on the ends of both sides a U - shaped bracket and drill and tap holes and screw them to the ends of the skate_bearing for extra ridgity so they dont flex apart from the rail .

    see what u think guys/
    Attached Thumbnails Attached Thumbnails Skate-o-rail-1.jpg   Skate-o-rail-2.jpg  

  13. #13
    Join Date
    Apr 2006
    Posts
    23
    Yep!

    Originally, I wanted to a "Y" configuration for bearings, somewhat similar to your two thumbnails, except upside down. It didn't dawn on me how to easily make it, so I went with four bearings to make things a little simpler to "manufacturer".

    Now that my machine is almost ready to start cutting, I can certainly revisit such things as the machine could make replacement parts!

  14. #14
    Join Date
    Apr 2008
    Posts
    1

    Unsupported linear rail flex fix?

    After looking at the design for these homemade linear bearing slides which looked to be nearly flawless, I wanted to add a suggestion for those who liked the design but were worried about the rail flexing with the use of this particular design. A couple of options crossed my mind here. One would be to secure the rail to an upsidedown mounted Piece of "T" steel or thin "U" steel that would be the length of the machine with a weld(assuming this material will take a weld and that there is a main support directly beneath the rail). The other option that came to mind was to tap the rail every 12 inches or so and run a few pins(welded) or bolts with steel sleeves or up through the main support of the unit. Welding may be the simpler of the two considering the fact that aligning multiple taps in a cylinder may be difficult for some. There looks to be plenty of clearance on the slides for either option.

  15. #15
    Join Date
    Apr 2006
    Posts
    23
    After many, many hours of using this homemade machine, I have to say that rail-flex really hasn't been an issue, even though during high speed use you can actually see the flex with the naked eye - a little. In fact, the machine isn't perfectly square in the vertical direction either. One corner is higher than the other three by about 1/16" or so.

    I've only used this for plasma cutting, as you know the torch shouldn't slide or touch the material being cut. So with the system not being square vertically, the height of the plasma gun varies from 1/16" to 1/8" - well within "spec" of my Miller Spectrum 375 to make fairly clean cuts in whatever I've fed it - assuming the material isn't super thick. 3/8" and less cuts fine, and clean enough for my purposes.

    I don't even clamp the metal I'm cutting down. No contact, no point.

    For slinging a dremel, or router around, where there would be resistance between the work material and the flex of the unsupported rails would be much higher, and the 1/2" diameter thin-wall tubing I used for the rails wouldn't cut it. And, why I don't do dremel/router work on this machine - I'd have to rebuild it.

  16. #16
    Join Date
    Jan 2009
    Posts
    3

    Links gone

    It looks like the links you posted originally are gone. Would you be able to put up some new links?

    Thanks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •