Revolutionizing 3D Printing: Prusa's Game-Changing Nextruder 

Really well made video tearing down the engineering. That's something most channels don't go as far in depth into, you have my respect

Are you still active??? I have been waiting for your extruder olympics to give some results to build a 'best' machine.

Nicecely done Olivier. Yes their extruder is well made indeed :) But still one issue remains i think... the straight teeth biting at constant distance in 90 degree causing the other source of artifacts we were seeing. I mitigated that issue with twirl so I wonder what coule be done here. Couple ideas are germinating in my head :)

There is also one additional bonus of having geared extruder, In case of very small extrusion rate due to small nozzle and layer height there can be artifacts caused by limited precision of steps in stepper motor. 1:10 gear ratio increase that precision 10 times, and yes I had some prints with 0,25mm nozzle on which stepper artifacts were visible.

This is a very good point, but I'm still interested in what'll be done with the strain-gage in the extruder for bed leveling, blockage detection, filament run out detection, and whatever else we manage to think of. All in all, I expect the Mk4 will be an important development for 3D printers...

Great point! Intrigued to see some real-world tests.

There's certainly a lot that is theoretically nice about the MK4, extruder included. I think it's still too early to really know if it lives up to the promise, especially considering Prusa hasn't released the input shaping firmware yet that will really let us push the print speeds and stress test the flow rates.

I would imagine that Prusa tested a large range of designs over the years after the MK3. They tend to aim for the Goldilocks zone where things are just right between performance, quality, and use. It's very interesting to see these advantage and how they might have selected this design. Thank you!

Although the bigger extruder gear and overall better design will definitely help with extrusion quality the best thing, I think is the strain gage. You could measure the force compere it with the requested force and therefore adjust the motor movement a bit to make it more consistent in extrusion.

Definitely looks like it could be a game changer. I am also interested to learn how this will work/perform with TPU. Really liked the dig into the mechanicals and the pros/cons of the currently used groved extruders. A most excellent video, thank you.

What happend, did you stop making videos?

Very interesting indeed but my biggest worries for the extruder is the curve in the filament it may limit types of tpu filaments. Most importantly brittle filaments like carbon fiber filaments as those don't like to be bent at too much of an angle before breaking. Can't wait for it to come out so we can see your testing!

Hi, have you considered having a look at belt driven extruder? I dont remember where i saw this but the way it function is that the motor drives belts on either side of the filament so the contact area can be increased according to the length of the belt.

Miss you and your quality content!

the only thing I have to complain or give feedback on the nextruder is: they should have made diagonal teeth on the drive gear, as the "imprint" it leaves on the filament will cause a repeating pattern of too much and too little plastic, while diagonal teeth would keep the deformation constant, it's a mostly easy fix (depending on manufacturing methods) and would eliminate the remaining variations to have a truly consistent extrusion

I was just watching your video on the Bondtech extruder with the layer issues and multi gear inaccuracies, and I thought "this is what the Nextruder is trying to eliminate". And here it is 😃

I can imagine this turning into a version of the debate seen on car enthusiast forums- running a single large turbocharger vs two smaller ones - only with extruder gears!

I recently went with the HGX Lite. It has dual 18mm helical concaved extruder gears. This video made me aware of the concaved issue and the difference is the HGX lite runs the extruder gears on needle bearings allowing the gears to slide on the shaft and always be centered. Prints are now amazing at any speed, VFA's are effectively gone. But even more noticeable is the extrusion consistency, for example i can run bridges at 250mm/s now that look and perform far better than before. The HGX lite is NOT perfect, and you have to be careful not to assemble it in a way that causes binging. Furthermore, i cleaned it out with WD-40 and compressed air, then assembled it to ensure no binding, ran it dry with no load to break it in, cleaned it out again!, then on final assembly added Super Lube (synthetic 30 weight with PTFE). I'm on about 500 hours of print time, with about half that enclosed at 50c. *printer runs at 200-325mm/s (0.4mm volcano nozzle) 5K acceleration

I agree with your reasoning. This could well be the breakthrough that extruders have been waiting for. Problems that were every time partially solved might well be solved all combined with this one. Thanks for the thorough analysis!

As soon as i saw the details of the nextruder i knew this video was coming. Well done and thxs for all the r&d you do for the community

Miss you bro! Hope to see your evaluation of this extruder one day.

i have a MK4 since day one and love the machine. prnit quality is extremly good. And besides what other people say, this printer is not slow at all with activated input shaper. THis printer earned a place in my 3d printing toolchain

A comparsion between the Nextruder and the Sovol all-metal planetary gear extruder from the SV06 Plus would be really interesting!

I think I'm more concerned with the gap between the drive gear and the entry into the filament guide than the drive gear fringe specifics.

Best graphics in the community. Thanks for your excellent work.

The sole purpose of the drive rollers is to ram the filament into the heated chamber where it melts and as the filament then acts as a hydraulic ram it is the pressure that the melted filament gets from the incoming filament that determines the flow.

My man just disappeared after this video

I do see a problem with older filament which has become brittle.. the bend might result in the filament breaking in the midst of roller bearings, clogging up the mechanism. I had that happen when I designed a rather similar (without planetary gears) drive for more contact.. it worked great, but was very sensitive to the usage of fresh versus older filament. Since I don't print a lot, my filament usually has some degree of staleness which mad it quite unusable.

Nice breakdown on the theory, but it would be interesting to see what, if any, difference this makes in real world printing.

Through the years I have always felt my Kossel mini with gear reduced stepper motor and simple gear and bearing compression system had just as good of results or better then I've gotten since. I still compare my best prints to prints off that machine from 2014.

What we also need is, consistent filament diameter, and volume. With the new large gear, see the possibility of a ribbon like filament.Not only the filament will be cheaper, it could also be more consistent in volume, made possible by the aspect ratio. It could pack more filament on reel, also could enable even more volume of extrusion. It might also be necessary to have a rectangular funnel shaped intake path on heated chamber. Just some thoughts .....

To be fair, I still think that lack of tolerances is the huge issue in terms of BMG. Genuine shaft was 4.96mm, while for this bearing arrangement it should be 5mm ±4um, so we're causing wear on the bearings. The ID of the gear isn't reamed either from basic visual inspection, so they probably don't have great tolerances here either. In my opinion just proper shaft OD and gear ID (with respective tolerances) should be significantly better than what we have now on the market

Very interesting to have this analysis ! Can't wait for the result !

Thank you very much for breaking these components down for us and giving your analysis. I wish you had pointed out that the only thing that's truly new here to even commercially manufactured extruders is the dual idler bearings, but that's me nitpicking.

Thanks for the insight... I have to start designing a new extruder now that I had a stable design... Excellent analysis and video as per usual.

it is precise enough that we can return to the conception of remore motors, and extruder gears driven by flexible shaft ? I miss those

Maybe that is why Ultimaker, until today, uses this kind of system.

Having seen your videos in the past the first thing I thought seeing the nextruder reveal was your video series.

Very good explanation… I always knew the print quality on the mk2s was better but couldn’t explain it.

wouldn't the grooves pushed into the filament create a much larger distortion in filament volume? You would be pushing filament volume equal to the displacement of the grooves up/down and partially to the sides. Assuming most of the plastic is pushed up and down due to where the force is applied, You would get 2x he differences in volume displacement from min to max groove depth. (ie. min = filament - groove volume, while max = filament + groove volume) If the groove displaces 1% of the filament from either side in regular printers, you would get a 4% delta between grooved filament and no groove. In this case, it may be slightly less since it only produces 1 groove, but they will be deeper/wider at their full depth due to the straight cuts. I have speculated for a long time, that ideally you would want a multi roller textured silicone feed system, to increase grip while preventing any change in filament shape. It may even be possible to use 2 "belts" pushing against the filament to feed it into the hotend. Another improvement could be to increase the length and thermal mass of the extruder barrel. This would allow a more gradual heating with a reduction in temperature fluctuation. Uneven heating of the plastic can cause thermal stresses and uneven thermal expansion/shrinking. This could lead to warping.

Yeah when i saw that i was like "wow that's so right and obvious in hindsight, why haven't we been building extruders like that all along?"

That massive gap underneath the gear looks like a great place to store my flexibles lol.

Nothing but problems with my last Prusa, never buying another one of their products again

This really does look interesting! Also… whatever happened to the hotend Olympics?? I never saw a results video.

really good video, full of nice information and well illustrated examples

How about comparing other big wheel solutions like Orbiter and LGX? A friend from the Voron community did some experiments and subjectively the nicest prints were from Orbiter 2.0 right after LGX Lite and RIDGA, the worst were the original BMG wheels and it didn't matter what extruder he put them in. The strange thing is that LGX Lite turned out worse than Orbiter even though it has bigger wheels.

I'm sold.

Amazing! Keep it up MirageC!!

Hy man! how is going with you? everything ok?

Bro this channel could of been a great channel, especially the hotend Olympics but all stopped... Why?

Proper printing’s timing belt extruder is revolutionary

Nice video! Though you missed one thing - the two idler bearings are not flat, they have a groove in the middle to keep the filament centered. Which means that manufacturing inconsistencies along the width of the gear are even less relevant since the filament cannot move left or right at all.

With all the potential issues in the extruder I really like prusas solution

Yes, the nextruder seems a big step forward. The bigger gear alone definitly is an improvement, as I saw on my zorbiter (just a little bigger) and my TBG light. The new concept with two flat idler bearings might improve even a bit more - but I wonder how much, as with the heatbreak integrated in the nozzle reaching up extremly near to the gears, there is very smal room for filament moving sideways anyways. Also the new nozzle system seems a good improvement. While you can get simular concepts from china (Trianglelabs CHC), the concept of changing nozzle including heater with just disconnecting two easy to reach connectors and looseing two thumbscrews is quite interesting. Esp. as you not only can use integrated nozzles but also a nozzle adaptor (heatbreak) to use special nozzles in v6 size. I hope there might be a version with longer meltzone for more volume flow (for prints with >0.8mm nozzle) And of course, the load cell always kills the need for z calibration. I hope I get my mk4 upgrade soon.

Very High quality content Brother!

You're wrong about the purpose of the 2 planetary gears. They simply support the shaft on the stepper motor, putting the shaft in double shear. Having a shaft in single shear puts a lot of wear on the bearing in the motor and will cause it to wear out prematurely. Supporting the shaft on opposite sides of the load, There's no movement in the rotational axis of the drive motor. There is no reduction happening because the Large drive gear is still mounted directly on the shaft of the motor. The Motor still drives it directly. The 2 planetary gears are nothing but a support on the other side of the filament.

I like Prusa's solution. Larger gear driven by a planetary gearbox. makes sense.

I would love to see a break down of the LGX Lite. Great video as always, thank you for the great info!

I wonder if the grooves indentation on the plastic also contribute to some inconsistency of extrusion. In that case the larger diameter with more smaller grooves is better. But perhaps there could be other ways to improve traction, like multiple stages of gears (with very small indentation)

Great analysis, as always 👍When can i get it on my non Prusa printer? 😅

Straight gear and planetary box- sounds like my Makergear from 8 years ago!

I have learned why crealitys next-Gen printers dont extrude the most consistent and besides there Top of the line machines are really hard to get that perfect first layer the extruder gears that make contact dont have a curved grove its straight cut but is still the same size gear which causes either Way too loose grip or Air Bubbles in your hotened garunteed from my 500hrs of testing with different tensions on my Ender 5 S1

The legend says he's still waiting for Sanjay's reply on Nextruder.

Fantastic! Subbed!

Hi man, I really appreciate your videos, I've a sprite extruder, which has the same issue in a way, so I'm planning to make an alternative arm (the one with the second driven gear), so I'm making an arm with 2 bearings to test this concept, I'll try to bring it as an alternative for the sprite extruder. Not that you're interested, but I'll keep you posted :)

It’s very interesting, but weight seems like an afterthought. Large metal gears seem like the antithesis of what you made with the hextruder.

Excellent, and well done. I only wish that the Nextruder wasn't tied to platforms which won't be able to fully exploit it -- the MK4 is held back by the old i3 kinematics, the XL is just pure unobtainium. As much as I would rather support Prusa for their innovation, it looks like we'll have to wait and see what someone else can do with a cloned Nextruder on a better platform.

Does anyone know if the stratasis DC motor extruder is patented? It's also quite simple and works very well

The next logical step is a live filament diameter check, which would re-calculate feed rates as needed. This is a good method in removing a few variables from filament extrusion. What is maximum feed rate this this extruder? Is the stepper spinning 10x than before? I print a lot with 0.8 and 1.2mm nozzles and have maxed out various extruders, limiting my production. I hope this isn't going to make the stepper spin too fast and create a skip.

yeah never thought of the dual hub thing being an issue. but it deffinatly would explain the almost z-wobble type behavior even though i have a belted z.

I just love this video.

Very good analysis! To modify the current system could you use two plain ball bearings with a V-grove between them as the idler? Or a wider bearing with a V-Groove machined in the middle? This could keep the filament centered better and distribute the pressure over a wider range. Thoughts?

When new Video

Interesting analysis!

Thank you for actually doing calculations! All too rare on 3DPtube

This feels like a really DIY friendly design, improved extrusion with less manufacturing tolerances. Any plans on including this on a new extruder? It seems plausible that a precise drive wheel can be sourced from some existing other industry. Or should I get a Patreon subscription to find out about that? ;p

They didn't explain how they're going to retract the filament. To retract at 80mm/s, the stepper motor will have to spin 10 times faster than if it was a direct drive. That would be 800mm/s... At those speeds, it loses all the torque

Nice work. I still perform dual gear system. It would be good to have bigger pulleys but both driven. But this is just my opinion, it would be good to measure it somehow..

soooo.. off topic from everyone else, but where did you find the octopus hot end fan ducts STLs that are in your video?

Great analysis

Would a helical gear not be better pulling the filament as opposed to the straight cut pushing?, would that not create a more consistent pressure in the nozzle?

My mk4 just stopped due to z axis calibration and is a nightmare to fix

Interesting analysis, thanks

I'm waiting for Nextruder-ish Aliexpress versions or something similar from Bondtech

Prusa should give you credit for coming up with their ideas..you had been the origin.

Nice info, thank you for sharing it, keep it up :)

excellent show thanks

Prusa slicer runs at 95% flow rate lol as default 😊

While the large surface on the filament and the efficient planetary gear are great choices, they come with one downside: They are big, resulting in a long filament path to the nozzle. Depending on the flexibility of the filament, we might see some improvements with this type of gear, while we see downsides due to filament compression on the path to the nozzle. The resulting question is: Do we really need large gears? Do we really need maximum tracktion? Would a small gear on one side and a bearing on the other side not do the trick? My assumption: Unless we work with extreme pressure in the Nozzle (speedprinting), all of that might just be unnecessary. An updated Hextrudort with a straight gear and bearing might just provide the best of both worlds.

Hi. I admire your perseverance in investigating the causes of the heterogeneity of extrusion. You've done a lot of work! I have an immodest request for you. I'm developing an extruder drive that you should like. I called it SL_ST EXTRUDERRR. Can you take a look at it?😅

A question is how heavy this thing is going to be as there're quite massive metal gears there?

I can’t wait for you to review one either. I tried to double subscribe but RU-vid wouldn’t let me!

Beautiful video!. I do have a few questions. In your comparison of hobbed vs straight vs large straight, are you ignoring that in dual drive systems there are 2 gears in contact? that would mean 2 * 0.749 = 1.48 ^2 vs the single large gear of 0.964 mm^2 Why not compare to Bondtechs LGX series? with its 18mm diameter and dual drive this should be even greater. Finally the hypothesis of the filament walking is certainly possible, but I think the curved surfaces will self-center the filament. and have a proper filament guide will also help there. did you have evidence of the filament walking?

Cool tech, but its pretty disappointing that all prusa did was add an extra idler arm instead of making a geared extruder.

tri has the similar large gear extruder maybe search for tbg-lite

How would the filament drive the teethed gears? The larger gear has a stepper motor, and I don’t see the filament being pulled during printing being able to overcome the stepper motor.

I counted 14, 56 and 126 teeth, which is a 9 fold reduction. Not 10.😊

@5:10 yeah, still waiting for the hotend comparison...

Why isn’t this extruder put in core XY 3D printers?

Without modding it so the idlers are cupped to replace the cup of the hobbed drive gear, I think this is a non starter for flexible filaments. I don't see how it could constrain the filament not to just shoot out the side.