RepRap Wally Can Print Larger Versions of Itself

SCARA based 3D printers seem to be all the rage these days, and with good reason. This RepRap Wally doesn’t use any linear rods or timing belts — in fact, it can even print larger versions of itself with each iteration! Well, minus the electronics of course.

It was first spotted out in the wild at the NYC Makerfaire, and looks to be a pretty slick design. Using fully 3D printed limbs, the steppers move the arms using a fishing line. To reduce the load on the joints, a bowden extruder is also used. The really cool part of this is the z-axis, it uses a 4-bar linkage to stay level, but because of this, it also moves along an arc in the y-axis as it raises or lowers. This is accounted for in the firmware — otherwise you’d have some rather interesting curved prints!

Stick around after the break to see it in action, it’s a nice change to watch from the standard gantry style printers.

Looking for more SCARA based printers? We recently shared a giant industrial SCARA robot that has been converted to a 3D printer!

[via Reddit]

10 thoughts on “RepRap Wally Can Print Larger Versions of Itself

  1. Instead of compensating for the curved motion of the 4 bar linkage for the z axis, they could’ve used a Peaucellier-Lipkin linkage.

    1. While that may seem like a good idea, the 4 bar linkage has way fewer joints and linkages than the Peaucellier-Lipkin linkage. Since most of the slop comes from the joints, fewer joints means less slop, which means tighter tolerances. Accounting for the lateral motion is done in software, which doesn’t suffer from slop. Adding the software translation is also a lot cheaper than printing more linkages for every printer. The 4 bar linkages is cheaper and more accurate. The only downside is the development of the software translation.

  2. Sarrius linkages seem to be one of those things sounds like an awesome idea for reprap, but I’ve not seen a working one in a printer yet.

    I’ve thought many a time about how cool dual arm scaras are for reprap, though I want to put the motors on a rising platform, not the bed, to do away with the need of a back board.

    Bowden extruders are the reason I don’t have a delta, so it would be good to have arms strong enough to support a direct drive extruder.

    I’ve also wondered if those brushless gimbal motors would be good to drive the arms. Though they’re the same price as steppers the electronics to drive them could work out a lot cheaper.

      1. I saw that, Pretty cool (I’m same commenter as above).
        I did wonder about slop. I guess the 4 bar linkage does work well, though it does reduce the Y range. I wonder about the slop with the bed being just pulled up by string and down by gravity? I guess if you keep the acceleration low it will be fine…

        1. The solution is obvious: http://www.dumpt.com/img/viewer.php?file=auq6si8h295x8r2dyiz1.jpg

          If you connect both halves by string and pulley, they act as counterbalances so that when the upper half goes down, it pulls the lower half up, and vice versa. Some clever routing of the string shall be required to eliminate the differences in geometry so that they both move at the same rate.

          Maybe extend a lever backwards from the upper arm and pull a string from that to the lower arm.

          1. Wally creator hear. I am looking at using a variant of the Sarrus linkage. http://www.youtube.com/watch?v=otHNw_ApOJ4
            I will try this with 4 links before I do it with all 8 links shown. I have my fingers crossed.

            @Dax: I really like the counter balance idea. That opens up some interesting possibilities. We don’t need straight-line motion. We need relative straight-line motion.

          2. I was also thinking the upper platform can simply be a bracket that houses a simple passive scara-arm to hold the tool level. The movement itself would be accomplished by strings or belts around the perimeter based on this design http://www.corexy.com/theory.html for cartesian movement, so there would be no need for computational transformation.

            Plus, the arm can be loaded with sensors to detect the position for feedback operation for additional precision and motion control. If you can get a feedback loop going, you can easily account for slight errors in geometry.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s