A High Speed, Infinite Volume 3D Printer

One of the most interesting developments in 3D printing in recent memory is the infinite build volume printer. Instead of a static bed, this type of printer uses a conveyor belt and a hotend set at an angle to produce parts that can be infinitely long in one axis, provided you have the plastic and electricity. For this year’s Hackaday Prize, [inven2main] is exploring the infinite build volume design, but putting a new spin on it. This is a printer with a conveyor belt and a SCARA arm. The goal of this project is to build a printer with a small footprint, huge build volume, no expensive rails or frames, and a low part count. It is the most capable 3D printer you can imagine using a minimal amount of parts.

Most of the documentation for this build is hanging around on the RepRap forums, but the bulk of the work is already done. The first half of this build — the SCARA arm — is well-traveled territory for the RepRap community, and where there’s some fancy math and kinematics going on, there’s nothing too far out of the ordinary. The real trick here is combining a SCARA arm with a conveyor belt to give the project an infinite build volume. The proof of concept works, using a conveyor belt manufactured out of blue painter’s tape. These conveyor belt printers are new, and the bed technology isn’t quite there, but improvements are sure to come. Improvements will also be found in putting a small crown on the rollers to keep the belt centered.

All the files for this printer are available on the Gits, and there are already a few videos of this printer working. You can check those out here.

22 thoughts on “A High Speed, Infinite Volume 3D Printer

  1. You know, that’s the problem with 3D printing…
    Just when I think I’m ready to take the plunge and buy one, something new comes out that could render existing ones obsolete.

      1. I tend to think FDM-style 3D printing has reached a plateau. The last major technological change that really effected quality vs performance was the widespread adoption of PLA filament (which fairly consistently gives good parts even without a heated bed). It has been about 6 years since then, and so the technology has been fairly commoditized now. You can get a small-build-volume 3D printer for $150 to $300. This one (which I haven’t used) in particular is cheap especially given the heated bed, ability to print headless (with a display), and with automated bed leveling:
        https://www.monoprice.com/product?p_id=21666&gclid=Cj0KCQjwnfLVBRCxARIsAPvl82E94CkJJj79bGg6-7BHZy6f2qtA-xSy8h39ZAmlCp5ppP_MrXAU4fIaAuN3EALw_wcB
        Honestly, I doubt it’ll get significantly cheaper than that. (although get rid of the display and heated bed and use USB C for power, and you may do better, but there you’re sacrificing features).

        I bought a Printrbot Simple Metal about 2 or 3 years ago, and the pricing on that has remained about the same in that timeframe.

        VR/AR is about 3-4 years behind 3D printing, so there’s still a fairly rapid improvement in technology (although slower than one might have expected 2 or 3 years ago).

        1. It’s like saying “everything that can be invented has been invented.” I think there will still be some breakthroughs. Just look at MRRF, there were many cool things there. Like new metal infused filaments (well, I still have a bronzefill sample from last year, so that’s not that innnovative) or head-changing. There will be innovations, we just don’t know about them yet.

        2. Not really.

          You could argue that FDM has always been the same, and never progressed. Using plastics like PLA sure make things easier, but I’ve had machines designed exclusively for ABS that I’ve used exclusively for PLA, so in the end, is that an advance at all?

          If you look at things like octoprint, or power loss print saving, the improved stepper drivers or boards going into machines, magnetic beds or PEI powder coating, multi-matieral solutions like Prusa’s MM systems with bowdens or the new filament selectors that still use only one nozzle, or ruby nozzles or quick switch nozzles, you could argue that FDM has been moving at light speed and will never stop.

          Just because a concept is still recognizable after a decade or two doesn’t mean it’s flatlined.

          1. I didn’t mean to imply that FDM has flatlined entirely in terms of feature set. But the rate of improvement vs price for entry-level machines isn’t likely to change a lot in the near-term. $150 for a printer with automated bed leveling, a display, and a heated bed is unlikely to be dramatically beaten without sacrificing some of those useful features.

            There will always be new features added, but I think the beginner is not in danger of being rapidly left behind by getting a decent entry-level machine right now.

            There will be another wave of broad improvement as the heated build chamber patent expires (sometime in the next year or two), although people have already found ways around that: https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20170000214.pdf

            …but PLA in many ways made headed build chambers unnecessary for entry-level systems. That’s why I consider the widespread adoption of PLA a major improvement that allowed a step-change in quality across the ecosystem. By eliminating or reducing requirements for a heated enclosure, PLA not only side-stepped the patent, but also enabled 3D printers far cheaper and safer and with less environmental hazard due to ABS printing fumes. It was a major, major change.

    1. I know what you mean! But don’t worry, most of the new developments only apply to niche applications. Just get a cheap one and start experimenting!

      And you can always retrofit if you really want to add new functionality. After all, you’ll have all the steppers, controllers, belts, rods, bearings, etc. to build off of. My printer is constantly in flux. Thinking about adding a second extruder / hot end soon, as well as changing out the motherboard and stepper drivers for something quieter and more accurate.

      Converting to a conveyor bed would be tricky, but doable. I haven’t had need for a super-long print yet myself. And there’s technical complications, such as supporting the printed mass as it gets longer. Not too difficult if you can set up a row of rollers on the output side, but obviously the theoretically infinite print volume becomes quite finite and well-defined based on how long the output table is, before the print starts to lever off of the print bed and lose adhesion. That would produce some interesting failed prints!

    2. By that logic, you would never buy a computer, as they always get better. :)

      These printer-belts are very, *very* interesting, but are still *years* away from practical use. There are present examples that work … mostly. Kind’a. With some exceptions. You probably want to stay away for a few years, until the bugs are worked out. (That said, I will likely buy a PrintrBelt off Brook in the next year, but not just yet.)

      Just learning to deal with the quirks in the present generation of 3D printers is challenge enough. There are a lot of quirks, a lot of folklore (much dubious), and enough odd behaviors to keep you occupied for quite some time.

  2. Next step, someone makes one of these but uses an omnidirectional conveyor belt instead of a linear one.

    I wonder how long until someone puts a 3D printer in open space and it truly has virtually unlimited build area. Less potential for sagging upto a point.
    Maybe the real solution would be to put the extruder head upsidedown in a fixed position. Then suspend the print above it with helium balloons. :D haha

    1. Get yourself a robot arm from an old abandoned car factory, bolt your 3D printer where it’s tool goes. Take the bed off. Viola!

      If you want more volume, load the arm onto a forklift, a self-driving one ideally.

  3. The problem with putting a crown on the rollers is that the belt will conform to the rollers and display the crown shape in the area where you want the belt to be laying flat, a better idea if you want to try and force belt tracking would be to add a v-belt to the inside of the belt and matching grooves in the rollers..

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.