Josef Prusa: Multi Material Extruders For Amazing Color Prints

The Prusa i3 Mk 2 is the hotness in consumer-grade, quality 3D printing right now. And things just keep getting more interesting. We caught up with Josef Průša at Maker Faire Bay Area this weekend to see the multi-material extruder in its final form. It’s an upgrade to the Mk 2 which allows a single hot end to print in four different materials, be it different colors or different types of filament.

A Bit of History

Final version of the Prusa i3 Mk2 with multi-material add on

Jo has a reputation for turning out some of the most clever and dependable design improvements to 3D printers. Just one year ago he was at Maker Faire showing off the new version of the popular i3 printer, called the Mk2. He followed that up in September with an announcement for the multi-material upgrade which he had on exhibit for the New York Maker Faire last October. Check out Brian Benchoff’s coverage for in-depth details about how the Multi-Material extruder works.

A few details that went into manufacturing delayed delivery. But customers who pre-ordered the upgrade are now receiving their units. Anyone ordering now can expect delivery starting as soon as August.

Why it Works

In the past, common approaches for multi-material printing used a separate hot end for each material. But the big flaw in that design is the added complexity. With more than one hotend you need extremely precise calibration so that the nozzle heights (Z axis) are exactly the same and the X/Y alignment between the heads match.

Prusa Research sidesteps this issue by using a single head. Four Bowden tubes all connect in a four-way splitter on the print head. When it’s time to switch materials the head travels to another part of the bed and purges the remains of the previous material before returning to the print with the new filament.

As you can see, the results are spectacular. The alignment of different materials is perfect. Add to this the already great resolution and printing capabilities of the Mk2 and you have another leap forward in affordable, high-quality printing. The printer itself starts at $699 and the multi-material upgrade is $299.

61 thoughts on “Josef Prusa: Multi Material Extruders For Amazing Color Prints

      1. You can’t mix, it’s one filament at a time to the print head. Didn’t this already get coverage a few months ago? I was also under the impression that it’s multi material capabilities were limited to those with a similar extrusion temperature.

      2. From the way they’ve done this, it seems like with a bit of tweaking you’d just have to send different filaments through at the same time. You’d have to adjust the feeding to keep the pressure OK. Not an expert, but it seems like just a case of “some engineering required” rather than reinventing the Black-Hole Reverse-Quantum Drive.

    1. No, what we need is for Prusa to stop messing around and do something about the 7 week backlog he STILL HAS on orders for his 3d printer, 9 months after it was released.

  1. Perhaps if there could be a hot chamber for mixing pigments in before the nozzle, we could have one single blank (i.e. transparent clear) filament, and add pigments as needed to tweak transparency and colour. With mixing controlled, we could get smooth transitions, or make a stop, purge, load new colour mix and continue.

    1. An interesting idea but it would require a fundamental redesign of the extruder’s functionality and may result in structurally weaker prints. It seems worth investigating though.

          1. Hmm, white filament with Cyan/Magenta/Yellow/Black sharpies surrounding it. Each on a servo that can optionally touch or not touch. Then, PWM each channel to get everything you could ever want.

    2. Actually, there has been some action with adding color to white filament by pressing Sharpies against it, just before it goes into the extruder. Now it’s up to someone to take CMYK sharpies and make some kind of carriage that will color the filament on demand.

      1. That’s basically just a solenoid and a bracket. Surprising that so much has been done to advance 3D printing, but this hack has been known about for however long and nobody’s worked on perfecting it.

        Actually as far as the “bracket” goes you could 3D print that bit. And the solenoid’s core. So, wire, then.

    1. Consider that most everything that you own that’s made out of plastic was injection molded. And injection molding inevitably leaves behind a significant mass of sprues and flash after trimming. Not to mention the energy consumed to keep the die and extruder at temperature for days on end.

      1. injection molding machines work 24/7 and whatever is not part of product goes into grinder and back into machine. There are exception to what materials can be reused however.

        1. Are you saying there could be a few unfortunate bits of plastic that always end up in the sprue in spite of having been through the cycle 14236994 times…? Dang, that’s harsh…

    1. Lots of material waste, not being truly multi material is the reason why most other manufacturers (including us “Ultimaker”) are not doing this.

      For example, you could never do this with PVA/Nylon (which is a great combination, as PVA sticks well to nylon and the other way around). But the PVA would burn up and clog your nozzle at Nylon temperatures.

      True dual hotend systems just provide a lot more flexibility. For example using a 0.4 and 0.8 nozzle in the same print. Using a specialized nozzle for a certain material (like we do for PVA)

      1. So, with a single point of entry for the filament, what about a way to rotate different hot ends e.g. have hot ends at different temps/sizes/used for different materials?

        1. That is not so easy to make mechanically. Twisting wires become a thing, especially in rotating set ups. Robot cables are expensive. Also, the fans on the side of the hot end complicate things.

    2. Would it be possible to print in a hidden area while the transfer takes place? That way there wouldn’t be so much wasted material as the colour change filament would be dumped within the part instead.

      1. That makes sense. Infill instead of a dump area? Probably trickier Gcode generation, but if it’s calculating for material changes anyway that shouldn’t be that difficult. If it’s done as solid infill, that ought to clear it pretty well.

        Seems obvious to me, but if smarter people than me haven’t implemented it, there must be a reason…

        1. I’ve seen it mentioned in the context of a slicer developer who hadn’t gotten around to implementing it yet.

          In 3D printing there are lots of things people have though up, but haven’t implemented yet. (Non-horizontal layers for better bridging. Variable infill. Optimization for specified required load forces. etc.)

    3. What it needs is the same thing inkjet printers have, a prime/purge station and wiper. Have a blade at one end of the gantry to cut off the purging, with a chute to direct the waste away to a catch tank. No need for a prime/purge tower to waste a lot of material keeping up with the build height.

      Most inkjet printers have an absorbent pad in the bottom of their case. It’s usually quite large, intended to last the lifetime of the printer. Some of the higher end HPs had an easily removable setup with a small pad. The whole thing could be hosed off with hot water in a sink to wash out the pad.

      Such a setup on a FDM printer would be far less messy.

  2. Work out the problems with mixing, adhesion, and on/off control, and get dot matrix 3D printing (the “dots” being little blobs of varying-color plastic). Maybe about a googolplex times easier said than done.

  3. It’s cool that he can print Blistex I guess. I mean I don’t have face herps or nothing, but its a neat trick. But isn’t it like trademarked or something? I mean wouldn’t P&G have a beef with you just printing tubes of their product?

  4. Only problem is it can only print properly at 3 quite large resolutions once it’s been converted to the MM upgrade, and the waste tower is often bigger than the model you’re printing! Also as it’s a bowden setup printing with flex maybe very problematic!

  5. I wonder if there is a plan to figure out less waste. I like the idea of printer knowing when its going to need to change colors and use the amount for the infill or support instead of just wipe on a separate area. I can’t wait to see if you guys figure this out because your guys printer is pretty awesome.

Leave a Reply to MaxCancel 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.