Printable X/Y Linear Rail

by PacManFan Jun 10, 2011
Download All Files

Thing Apps Enabled

Please Login to Comment

Comments deleted.

and.. I just wanted you to know...

the Rack is a bit wider than the groove in the "tube" so when i printed it i could not fit the rack inside the groove...

I printed a 250mm tube so.. i guess i have to modify the rack and print new ones, so it fits the tube rather than the other way around.

Great project.. THANKS

I just wonder.
what are the purpuse of these "lips" on the small gear? https://dl.dropboxusercontent.com/u/21839627/smallgear.jpg
Do they have to be there or are the just a product of a bit fuzzy openscad code?

how do you mount the small gear to the motor axle? I can`t see any set screw or anything on it.

I would love a Nema 23 version and with some kind of set screws on the small gear for this... then I would totally use this for a small CNC mill/engraver.

Everyone who's been asking - I have uploaded the rack.stl file, Please give it a print and let me know what you think!

I'm playing with rack-pinion idea right now and start to like them and would try to implement it in my current project, could you add the rack too among the files ? :)

Brilliant work!
Do you print the rack with the rail or print them (horizontally) and glue them together later?


can you put on the toothed rails as stl? :)

interesting, where's the Rack.stl?

How about printing the rails directly on ABS pipe?

It would be faster, cheaper, stronger and an easy machine to build.

See the picture for explanation.

One thing I'm going to be posting soon is the scad and stl files for an entire printer design I've been woking on - the "MultiRep V1". It uses rails similar to the ones here, but the shuttles are not self propelled. It uses the printable leadscrew design I posted before. It's a basic A-frame design like the Mendal. I actually started printing the Y-Axis of it last night. I can't wait to show everyone. I also have a great design/print for a planetary geared driven peristaltic pump that's going to be used as a paste/ceramic extruder.

Interesting concept. One variation of this is to have 2 counter-rotating tubes, each tube is a threaded rod similar to the ones I posted, but thicker to support the weight. The shuttle then is both propelled and supported by the threaded tubes. The biggest concern would be to ensure the tubes are thick enough not to sag for the entire length.

First axis done, works well. Now on to the next!

Any thoughts on a z axis? There are quite a few on thingiverse already, so I may try a couple.

Could you just let the motor be stationary and have the rail move up and down.

I'm currently working on a printed leadscrew design for a z axis. I'll post very soon.

Is there any reason why one of these rails couldn't be mounted vertically?

Good question, Honestly, I haven't tried to use these vertically. The biggest concern would be that the motors don't have enough torque to move up and down. Another concern would be that once the motors enable goes to zero(power off), gravity will pull the shuttles down. Give it a try and find out.

For the z Axis you could modify the shuttle to have a worm gear from the stepper to a half worm half herringbone gear that goes to the rack or some such variation. That would take care of the torque issue on the stepper due to weight of the z axis and gravity

I got some slipping when I tilted my rig 90 degrees (I'm lazy) but after quickly tightening the big gear to the rail it works.

I just realized, with the first axis I set up I've inadvertently matched the x dimension of the cupcake. Since the plan was to make it square I guess it's pretty easy to ballpark my final size now :)

You should add a V groove to the rail so a groove bearing can be used instead of 2 608 bearings on either side. See this kickstarter project for a pic and video about grooved linear rails.


My first attempt was with a v-groove bearing. The v-groove was easy enough to print, the v-bearing was not.

Weren't people making the v-bearing parts in 2 halves, and gluing them together? Is that too much of a hack?

You should get Flattr

Looks great cant wait to see more.

I was wondering where i can get the file for the straight herringbone rail that interlocks with the tube. This whole project looks great!

Hey, what is the purpose of the "arrow" gears? Does it give more stability? I'm not much of an expert on gears, but I would like to use this for a RepStrap and I was just wondering if it had any purpose besides looking cool : )

I'm using double-helical gearing for the gears and the rack. They are designed to reduce backlash when changing directions and keep the gears centered due to mirrored axial thrust. And yes, they look pretty cool.


So what are the 4 new holes on the carriage around the single bearing for?

They are general purpose mounting holes. That's how I connect the rail to the perpendicular shuttle. I'm glad to see someone else printing the design. Let me know how it goes for you.

Trying to clean up some sloppy print bits on the gears to get better meshing so I can turn up the Speed, other than that I'm thrilled!

Like the design a lot, I scaled the Tube portion @ 1/2 height to increase the number of interlocking parts. with the strength on this I'm not sure it was necessary.

I noticed on the X/Y gantry you made you put a stepper on each side, I'm thinking of trying just 1 and a free sliding carriage. Did you try that? If so, any tips?

also, if you really want to cut down on hardware, I've had rally good luck with building http://www.thingiverse.com/thing:4676http://www.thingiverse.com/thi... into all kinds of things. Build the inner part into one part as a single piece, then split the outer part in two screwed/snapped or even glued together. If there is enough support material it works great. You could end up trading a bearing vitamin for 2 screws and some BBs, cheaper and easier to come by.

Printable 608 Ball Bearing

I've printed a few of those out, they work really well. I really need to find some bb's that don't have a flat edge on them.

for those I kinda... cheat... I have a coilgun I built a while back that fires perfectly sized ball bearings, of which I have an awful lot... My friends have given m some plastic 'airsoft' pellets to try, I just haven't gotten around to it. Now I have a good reason :)

Excellent work, would love to see a video of the two 2 axis in operation. Are two motors needed for the Y axis? would it not be better to transfer the power through the center of the rail to the other shuttle via a printed/threaded rod axle?

you should get flattr

Looks great. Why the strange overshoots on the corners of the gear models? Does this somehow improve the printing.

No, that was just an error in the gear script. I'm going to re-write some of the gear generation code myself. The gears work fine with no perceptible backlash.

Very impressive. If you segment the rail design and design in puzzle-like interlocks, you could effectively accomplish a self-printable rail design. :D

That video is a perfect real life example of change blindness btw. If you look at the small gear it appears to be not turning but when you pause the video to compare it's position it obviously is... here's another video describing change blindness...



Its a perfect example of shutter sync.

The same phenomenom that makes car wheels on TV look like they are going backwards, slowing down then changing direction, all while the car moves in one direction and simply changes speed.

The same effect also makes scanlines visible on CRT screens shown on
camera. Amongst other things.

I see the car wheel effect in real life without a camera. This could be shutter sync but wouldn't that cause the gear to appear to be moving backwards or at the wrong speed? When I watch the video I can't see the small gear turning unless I stare directly at it.

It depends on a lot of factors, but mainly the frame rate that chewube is playing things back at. The video is classic shutter sync, I'm willing to put money on the small gear looking like its not moving even in single frames. If you watch at the ends of the rail you can see the small gear slow down.

I've never noticed anything comparible to shutter sync in real life except at very high RPM's. Not the sort of speeds car wheels normally go at, but I'm sure it depends a lot on the observer.

You can observe the shutter sync effect in real life at night on a bicycle; watch the spokes of your wheels while driving under the light of a fluorescent lamp.

wow, that runs very smoothly in the video! amazing work!

I think you should use this design for the body of the printer:


This is a snap design without screws , nuts , washers , threaded rods and etc.

I saw that. I actually designed a snap-together too last month but never posted it. I got halfway through printing it and decided it was taking way to damn long. I should post the plans for posterity, but I considered it a dead-end. I opted for a tube structure that could act as a rail and a support.

You can also use the chain from this design:


To make the motors stationary and design a fast printer like this:


Like the ultimake.

Caterpillator v1.1
by Olalla

this is quite an accomplishment. I can't wait to see the results of a printer design using it. You may want to try flipping it upside down so that the motor is stationary and the rail is what is moved, that way the weight of the motor is not carried. Either way the movement seems very smooth and quick so far. Another thing to play with would be an alternative to the skate bearings to marbles or something else that people have been experimenting with, one thing at a time I suppose... this is definitely progress!

Thank you, My next step is to use some of the printed 608 bearings, and replace the two back bolts with a plastic block for the rail bearings.

I had thought of fixing the shuttle in place and moving the rails to create actuators.
I would also like to try something like this design for a tripod:

What's the advantage of the "tripod" thing over a normal cartesian thing like this: http://www.youtube.com/watch?v=Cq5ouIPg-YUhttp://www.youtube.com/watch?v...

... well except for looking "cooler" ;)

I think the tripod design is more elegant and simple but perhaps a bit less sturdy and requires more complicated movement control. It requires three sets of identical parts, so not only are there fewer parts overall, there are fewer unique parts in the design, making it easier to assemble, repair, and make improvements. I think it's definitely the best way to go for anyone interested in taking the Gada prize. http://humanityplus.org/projects/gadaprize/http://humanityplus.org/projec...

I'm going to clean up the scad source and post it soon.

What did you use to design this? Have any source files?

I guess the gears could be improved. They don't look like the heringbones are involute.

Could you post the original files too please?