3-way Quick-fit Extruder and Colour Blending Nozzle

by RichRap Aug 23, 2012
Download All Files

Thing Apps Enabled

Please Login to Comment

I was thinking instead of using greg style extruder would it be also easier to use bowden style extruder? It would save some weight for the X axis and also might work for rostock/kossel printer

Wouldn't it be quite simple to build a nozzle that mixes the colors only just before the outlet ? this should provide much better mixing and would also help with very fast responding color changes. Of course it would be even better if there was some sort of tiny mixing chamber with some sort of mixing propeller...

How do you prevent the molten PLA out of one tube, to flow into one of the other tubes?

Was thinking about that too... But logically thinking, it´s quite simple. For the same reason the filament is not flowing back in a single element! (I think when the filament is pushed the force squeezes the soft, not yet molted part to thicken and it blocks the way back.

Soooo Cool. Would you be willing to sell me one with everything that I would need to get it to work with my Mendel?????

Thanks everyone, it's always nice to be featured.

Lets keep open source development going strong, lets continue to share, develop and innovate together. What at Awesome open community we have.

As always please make one, make it better, bigger.

All designers, hardware/software/firmware developers out there lets keep RepRap in front for leading 3D printing development! 

We need good multiple extruder support, AMF format support, multi-material support, colour support, so much more to do.


Are there people currently working on stl 2.0 makers and interpreters? 

As always, RichRap presents some awesome 3D printing tech ideas, while implementing them in amazing ways :)

Keep up the AWESOME!

Being "Featured" is WAY overdue. Very nice.

Totally agree... This "thing" is one of the coolest on thingiverse.

Well done RichRap! 

This is so awesome!!!

Incredible. =-O

You HAVE to use Red Green Blue filament in order to make EVERY color you want !


Is it easy to control the mixture ?

Yes, to do a good job of a real 3D colour (color) printer you will need CMYK + White and also clear and maybe even support material, so 7 feeds ideally.

Oooh upgrade the upgrade. I wonder if you could put 7 tubes in a Bowen type configuration.

Yes, Have a fixed print head machine and moving X/Y table then you could put lots of filament tubes with a short bowdens and have 7+ extruder's in a circle above the print head. It'll look massive!

Smaller 1mm or 1.3mm filaments would help when you get this many feeds.

And we still need a way to translate a colour map from a 3D object to the slicing software, STL can't handle that bit :(

So, it's not easy to control, too bad.

Good work anyway ! I didn't say it in my first comment !

Actually, you would want a cyan, yellow, magenta source. RGB is for additive coloring like lights. CMY (and also blacK) are for pigments and plastics.


Hey RichRap!

I've got a couple other links that you might like. Have you seen the RepRap Pro color mixing or the Indy Maker? I can't wait until the firmwarez support this stuff...



love your threeway extruder. I'm printing up a set now. I've got some extra NEMA14s sitting around...

Have you considered using a lobed cam system with microswitches to control your Sanguinololu board switching?

Hi Matt,

Thanks, I did follow the RepRap colour mixing project with great interest, I talk about it in the intro on my blog about the colour blender, it was the work of Adrian Bowyer and Myles Corbett that pushed me to restart my project and get it 'finished' to this point anyway.

I had not seen the indig
ogo colour 3D printer, what a massive beast! shame it didn't get funding.

It would be great to use just one motor and cams to enable and disable single or multiple feeds, Nophead was also thinking about a similar way to do it with gears, so you could really mix in different ratio's 'mechanically'

I did think about putting a micro-switches underneath the printer bed so as the platform moves about a different colour is enabled, you could print different coloured things at the same time by placing them on different areas of the bed or dual colour by spanning two switches - but it got dropped du
e to lack of time.

I hope you make one, I do love the iPhone cases, I have not tried insane blending and filament chopping/switching all at the same time, that would be spectacular!

Loving this dude!!


Fine and all to cite patent work, but, RE: figure 3, there's no auger in this design- it's free and clear to operate!

This kind of raises another question I've wondered about: how does patent law apply to open source thing information? Are the coordinates of some ABS goo somehow illegal if a drawing of it is at the USPTO (or other country's P[T]O)?

No doubt this will end in rants about DRM and the such- but I've
thought twice about posting Lego parts!!


"how does patent law apply to open source thing information"

You should not sell products under patent protection - using for own interests is not forbidden.

Originally LEGO parts are older than 20 years (the longest time for a patent), thats why you can sell copies.

Technically, even if you violate the patent for your own internal use, you are violating the patent and you may be ordered to stop. Of course, if you quietly use the technology, you probably won't get caught...

Hi RichRap,

I made an illustration that shows in Pic 1 the usage of a guitar wire (thinnest you can get) ,

need to be twisted in all 3 directions,

first I tested the mixing with special prepared half moon filaments (pic 1)

and this works. Its really important how "good" the guitar wire is twisted
gt; best near to a filled sphere,

(perhaps you need to glow the wire to make it "soft" for less brittle. My guitar wire was removable, pull it out only.

An option is the usage of bor chips.



Hi Kaipa,

I get what you were saying now, I see how that may look like it should work to mix up the plastic, but in practice I have yet to see anything like that actually do a job of passive mixing. I did try a few things like this, including splitting the PTFE tubes and adding baffles, the plastic usually
just passes all around anything you put in it's path and continues on it's way down and comes out as two separate colours 'toothpaste'

Adrian Bowyer and his team did an experiment with very fine Stainless steel mesh that's about the same as you propose and it did nothing at all to aid in mixing.

ut if it's worked for you, I would be very interested, do you have any pictures of the assembly and maybe a video of the extruded, mixed filament? would be great to see if this is doing a good passive job of mixing.

Hi RichRap,

I´didn´t worked on this idea since since januar 2011 when startasys published theire patent.
Before I thought I could get a patent for this, beeing earlier than Stratasys in one own developement (showing them repraper are faster)
I claimed my patent in December 2010, but startasys have an earlie
r priory date (23. June 2009).

My testings that time where about soft filaments to be mixed with harder once. Tests have been promising but not ready.
Today I would test the hot chamber filled with a simple screw of 2mm. (nearly the same as extruder screw)

Interesting, but I don't think that will mix the filament either, and may just block the nozzle.

There is not much point in patenting an idea that does not work or give the results you expect so a proof of concept would have told you if it was worth claiming a patent. So I'm not really sure why you stopped development?

I'm also not sure if you could get a patent anyway, there is a lot of pri
or art in this all over the net. Patents don't really protect you unless you have the almost unlimited funds to support them.
And if you are trying to get around someone's patent with an alternative design to sell, then again make sure you have funds to back it up.

Even if you don't infringe, a big
company can take you out of business, if you are a success and that's what they decide they want to do.

Finally, don't stop developing ideas and trying things out. We all need to move forwards. And if you decide something you do is great and needs patent protection, I wish you all the best.

Hi RichRap,

it´s my style since almost 20 years to claim patents for my ideas, because I cannot test and develope all of these ideas to end. Sadly not.
If a company is interested they may buy a licence or not.
Thats only business as usual.

As soon as i find literature about somebody invented the product be
fore me I do stop my activities. Should the inventor do this.

In our theme, you already know in theory that mixing with a multi-feed extruder should work . You dont need to test it. Its obvisius as mixing chocolate and vanilla souce but not mixing water and oil. The idea, design and prototype is a
different thing. This is a very good work from you.

A really important question, i would get an answer is: Why does Strataysy allows this company PP3PD from china selling a copy of it´s own prototyping machine (including heated environment) in europe, which is an existing international patent? Se
en last year at Euromold in Frankfurt.


A couple of years ago, the reprap- project had to wait until the main patent of FDM runs out and now Stratasys allows copies of theire best products???
In effect not b
ad for repraper. Isn´t it?

sorry I
´m on tour. will answer at saturday

Great Idea, Great realisation! but it
´s also a Stratasys Patent from 2009.

Hi and thanks, if you do decide to make one, I would recommend making the melt chamber as small as possible so colour changes happen quickly. this prototype has quite a large nozzle and chamber so it takes some material to push out the previous colour.

I tested the mixing with thin drilling chips inside the chamber, this works without stiring help from outside

Hi Kaipa,

I'm not sure I understand, do you have a photo or drawing? - have you drilled random indents inside the melt-chamber? or do you mean something else?

That patent actually does not apply. If you actually read the patent, it specifically states that the output of the compouder as they call it produces homogenous filament in the range of 1.3 to 2.5mm.

This device does not mix the components, does not extrude in the size range specified, and directly builds the model rather than producing filament. IANAL, but I believe that what RichRap has devised is not even close to being covered under the patent.

sorry, but he patent fits, its a typical infringement

important is the first claim 1,

"filament in the range of 1.3 to 2.5mm"
lt; this is an option of the more detailed claimes in the patent, not a must

"rather than producing filament"
lt; this is also an option not a must.


Hello Kaipa

Do you have a link to the patent?


Verry nice, great idea, even beter result. I love it. great blog site.

hope one day to have something like this on my Ultimaker.

That woudl be verry cool

Thanks, it would be great to see many feeds and one nozzle on an Ultimaker, Maybe another good reason to do a Bowden version ! :)

WOW! That - IS - Beautiful!

Those are some really cool looking results. I think the lack of color mixing is a good indicator that the hotend design is producing a laminar flow. I wonder if the simplest approach to getting a homogenous mix is to tweak the Length/Diameter ratio in the hotend so that flow is just in the range of being turbulent and then overpower that extra resistance with a beefy geared extruder setup.

We do need much more work on a real static mixing nozzle, it's certainly not as easy as one would first think, especially in a tiny hot-end.

Adrian's team used a rod/paddle driven by a DC motor and an O-ring seal to 'mix up' the contents of the nozzle before extruding. Good results, a bit bulky.

Wow, just wow.

And on your massive MendleMax of course. Any lesser printer would collapse under the load. :)


My Massive MendelMax still only has M8 smooth rods, so y a smaller (normal) Max should work well and support the weight better having shorter rods.

I did only loose 15mm of X travel on the Big MAX, that's what the Spacer 3 way RTP.stl is for BTW if anyone was wondering. it activates the end-stop
before the left hand extruder crashes into the Z axis rods and axis!

I hate when someone beats me to it.

Nice work!

Mine is currently unreleased and handle only two colors, but it may get obsolete before I even release it.


Thanks Laszlo, I started it at Christmas! almost didn't finish it.

For me the best 'Art' results are with two feeds on at a time and colour swapping the inactive 3rd one, then enabling it, switching colour on the other inactive one, repeat.

A dual one would be good - maybe work on a Bowden version, that was going to be my next version, but I'm finding it much ha
rder to work without a Lathe :(