SMALL GEAR Extruders are NOT VIABLE ANYMORE?
ฝัง
- เผยแพร่เมื่อ 20 พ.ค. 2024
- Get metal 3D printing and other professional services for your projects from today's video sponsor - PCBWay (5$ OFF for your first order) - pcbway.com/g/584K9t
👉🏻 In this video I am testing 8mm diameter filament gripping gears versus 18mm ones. A way bigger surface area should provide more consistent control and prevent filament slip and deformation on high-flow hotends resulting in a more consistent extrusion. I will show max flow rate comparisons, a microscope shot of how filament looks when heavily forced through the hotend, and compare print quality between BMG vs HGX Lite extruders.
🧡 Support my work at PATREON - / diyperspective
📋 RELATED ITEMS TO THE VIDEO (Amazon/Ali Affiliate)
- HGX (Oblique gears) Extruder s.click.aliexpress.com/e/_Dnb... or hs3dprinter.com/pages/hgx
- HGX Lite s.click.aliexpress.com/e/_Dmf...
- Orbiter V2 s.click.aliexpress.com/e/_Dn0...
- Digital Microscope LM407-Pro amzn.to/3YckE2O
- JAYO Meta PLA amzn.to/3PxwUGQ
- ANYCUBIC PLA amzn.to/3W3MmwE
- filalab PLA filalab.shop/en/shop/pla/
📷 VIDEO SHOT ON:
- Fuji X-S10 amzn.to/3Zar9BW
📢 OTHER MENTIONED THINGS:
- GJB Test Cube www.printables.com/model/2931...
- Helix Test Cube www.printables.com/model/6993...
- Cali Dragon www.printables.com/model/2091...
🕗 TIMESTAMPS:
00:00 - The problem
00:31 - The surface area
01:08 - Testing setup
01:46 - 8mm gears (BMG)
02:46 - 18mm gears (HGX Lite)
03:46 - 8mm vs 18mm gears print comparison
04:34 - CORRECTION: Amps/Phase is already Amps RMS value (no need to divide by 1.414)
05:35 - Angled teeth on 18mm gears
06:13 - Aluminum vs Plastic frame extruder
06:28 - Final thoughts
🔗 YOU CAN FOLLOW ME:
Twitter: / diyperspective
Instagram: / diyperspective
#3dprinting #filament #extruder - วิทยาศาสตร์และเทคโนโลยี
![[LIVE] : ONE ลุมพินี 63 | คู่เอก "ยอดภูผา vs โซเนอร์"](http://i.ytimg.com/vi/RGVmBsegKEI/mqdefault.jpg)
CORRECTION: Amps/phase is already the Amps RMS value! You don't need to divide by 1.414! Sorry for the mistake!
I always love the level of depth with your videos. They let the data tell the story and it is a great scientific approach
Thanks, try my best to not waste viewers' time and provide quick explanations with real-world results! :)
Im curious about the difference between the helical gears and straight gears for tpu. On my vzbot extruder the filament twirls its way out of the extruder, at least medium and softer materials
The only thing i know for sure is when i was making my own exturders (modded prusa i3 style) the M8 Hob bolt was out of stainless steel,, you wouldn't think something small like Steel vs Stainless steel would make a difference,,, but the difference was a very clean & sharp & deep cut griping teeth on the Stainless steel compared to the cut on the standard steel which always felt blunt and shallow in the teeth when you compared them side by side,, and it didn't matter how many adjustment i made on the lathe when cutting them,,, there was no stopping the Stainless steel griping power. when i got a nozzle jam it would compact the filament into a tiny ball between the Hob-bolt and Hotend before it would chew a hole into the side of the filament,,, the plain steel bolt you just had to yell at it and it would get scared and start slipping...
Thanks for the video!
I think, we're actually limited by the hotend flowrate, not the torque. For example, you don't have to increace stepper motor current when you use volcano instead of v6. Just because underextrusion is caused by insufficient power of heating transfer from heater block to filament inside.
Hotends' and extruders' performances are closely related. So we want the best of both worlds for the best results. You won't be using the Titan or BMG extruder on the Goliath, you know.
The ratio of hgx extruder (1:9.5) is calculated with 8t nema 14 or the 10t version?
Tnx
Is it time for Reprap to make the jump to servos? We seem to be getting there.
I wonder how high power servo bldc with a break would perform vs steppers.
You could improve your grip (and deformation on the filament) by regulating pressure on the springed lever. Also 24mm flow rate is 300 mm/s print speed, so should it bother me? But thanks to the test prints, maybe the wear on parts of bmg could be a reason of that difference in quality?
Really what intrigues me about this extruder is the size. The added flow rate/consistency is nice plus but really for most people your hotend going to be more of a limiting factor.
Your video is amazing! Looking forward to see your next video on comparison between std straight and helical gears. U might also interested to see what the difference between extrusion gear with lock nut and “CNC made extrusion gear unit with rod in ONE-Piece“. Keep your great work man~
Helical gears maybe are more beneficial for smaller gear extruders, as even straight teeth on big gears grip the filament great.
Great video nice and short to-the-point. This is the video that made me subscribe. Keep up the great work
Thanks, will do!
I wonder how layer adhesion will be affected with higher actual flow? Is the filament going because it is fully melted or because it's just being squeezed through?
I mean the max flow rates weren't that different between extruders, they capped more or less at the same point. Big gears just provided a more precise control with lower filament deformation.
Very interesting. I will eventually go with dual gear extruder. And now I know if I go big gears I may have to change stepper. Thank You for that information
great topic, thanks 👍
When i went for print speed, i ran into under extrusion due to slipping, but that went away when i raised the hot end temperature.
And here we go with the subscription button. Well earned. A treasure channel for people liking technical things.
Maybe add ABS as sidenote/example as well, so people focusing on it also get some information.
It may be worth testing the Sherpa as well, it performs a bit better than Orbiter even with the small gears.
i made a double gear mk8 and was the best extruder i ever used ( heavy a.f.) the new one does not even compares.
well, at least for my printing, i dont see the need for bigger gears, even tho i print at 200mm/s with 0.6 line width and 0.2 layer height - so around 23-24mm³/s. Stealthburner with Dragon & CHT Nozzle. for 0.12mm layer height prints i bump it up to 250mm/s. Thats with a heavily modified Endet 5 (Cast Aluminum bed, CoreXY Mod etc...)
Also running a Bambu P1S and that thing is rippin'.... up to 350mm/s for 0.12mm layer heights, which it can reach with the accels and jerk its set to...
The problem of large gear extruders is that they need more reduction as one turn of the gear pushes more filament length, so you have less overall reduction / torque. Comparing modern day large gear extruders with an ancient BMG shows that they have overall similar performance, despite the fact that the BMG has just a 3:1 reduction. If you compare them with something modern like a Sherpa mini that uses a 5:1 or optionally nearly 6:1 ratio, or even better an Ascender with 20:1 worm gear you'll see that the small BMG gears have still something to say about performance (as far as I know it's the standard choice for people that want to max out a Mosquito Magnum+ or a Goliath).
Also the friction/grip equation is F x f, where F in this case is the force pushing the gear, and f is the friction coefficient (same between both), there's not any surface in this equation. You see shallower cuts on large gears because you have more of them, but you'd need a firmer spring to have any advantage (and you'd see deeper cuts). The reason why small gear extruders can grind filament while large gear ones just makes the motor skip in my opinion is because the latter group generally lack the torque to do that, not because of a better grip. Up to this day I have never seen a concrete evidence that large gears have any real advantage. And I say that as an LGX owner! Maybe for flexibles
3:54 if light setup is exactly the same, the extrusion layer to layer volumetric consistency is actually hugely improved, see on verticals. This is actually the sort of quality improvement i'm specifically looking for. I'm not actually looking for speed at all, THIS is what i'm looking for.
The filament effectively turns into a timing belt. Great for retractions. Nice clean indents should be ideal. Anything where the gear has to pull itself loose from the filament would not. So I agree with the large gear concept.
May I suggest you to cut your Y scale on graph. It will spread datas to make it more easy to get the trend.
That being said, it seem that there is not a lot of effect so far, unless if you crank the flow a lot. Unfortunately, we don't have theses high flow results on the final graphic.
Overall, I'll give you an 'E+' for 'Effort'.
Well, I've just removed the thumbs down you put 2 months ago. I guess you were feeling bitchy that day?
If the graph was cut on the Y scale, you would complain that it gives illusion of larger effect isn't?
So, shut the fk up and think a little more next time before giving bad review and low value comment.
Thumbs up from Montréal
Sorry about that confusion, you are a great TH-camrs, keep doing your great content!
Hello I got the HGX lite 2 id like to ask what's the Step Value, Thank you in advance for your response🙏🏾
It should be the same as for the HGX Lite. I set mine - rotation_distance: 5.65968
But make sure to test if that works properly with your setup.
@@DIYPERSPECTIVE thank you so much! I couldn't find this information at anyplace on internet🙏🏾
Im iñusing the titan extruder , 1 gear , 3:1 ratio , volcano setpu ,printing at 150 mm/s at 4000 mm/seg^2 aceleration for all lines (suports perimeters,etc) ,and works pretty well in an ender3 setup with Klipper 😊
thanks man, I've been having extrusion issues and this will help diagnostics
Hopefully!
@@DIYPERSPECTIVE I really gotta switch to a full metal hotend tbh ;D
PTFE hotends just degrade way too fast. Back in the day, this was one of the first upgrades that I did on my printers. I just can't bother dealing with things that I need to constantly think about. :D
@@DIYPERSPECTIVE totally agree, the maintenance cost in terms of hours is just nuts
The current rating for stepper motors in datasheets is already RMS current.
Yeah, you are right, my mistake, thanks for the correction.
i like haldis products, way under appreciated. try their K1 HF hotend, its impressive design.
could you make a video where you upgrade a cheap kp3s pro
That's not the content I make here. I just test stuff and show the real world results.
@@DIYPERSPECTIVEunderstood thx alot for the reply though
why did you stop at 32mm3/s? Would've been interesting to see how far those could go, even if it's temporary and needs stepper active cooling to test out
that being said, 32mm3/s is already excellent. I'm just always flow rate capped
Even 32mm3/s is not realistic in real-world printing. You should check out this video - th-cam.com/video/XeMhQpcMrCA/w-d-xo.html I changed how I test flow rates as from my experience the blob method gives way too high flow rate results to real printing.
I'm putting together a 60a tpu printer and was thinking large gears would be my best bet. Will be interesting if you or I get to testing that first.
For the TPU you also need the filament path to be as close as possible to the filament gears so that it won't have a chance to "escape" elsewhere when under pressure.
@@DIYPERSPECTIVE Yes, this is an issue with 95a as well. I'm looking at the tringlelab tbg-air, as the integrated nozzle looks like it can be positioned right under the gears
@@1fareast14something like that would be ideal or a lgx shortcut or ace with the integrated heatsink. But please dont buy a copperhead or mosquito, just fit any random copper v6 or volcano. Slice engineering charges insane money to the point where a genuine v6 (remember, 65 pounds or around 75 dollars) seems like a bargain
Run the small stepper at a higher voltage??
Also they have 1.8a versions of that stepper
I am perfectly happy with my BMG, I don't chase the latest thing or have a desire to print superfast, quality is my goal and I am happy how I am
Any dent in the filament from the gear will directly result in variation in the printed surface. I want a drive that leaves no marks on the filament at all. Maybe eventually there will be a printer with an extruder that flows evenly and a Z axis that is accurate to 1% on 0.2 mm steps. It would seem possible using a laser interferometer or a calibrated capacitive sensor array often used in digital calipers to provide feedback to the Z axis drive. Until this is achieved 3D prints will continue to have fish scales and lousy vertical sides. With the emphasis on super cheap bed slingers or super fast core XY printers, accuracy has been largely ignored. Nice to see it starting to come to the forefront.
Maybe the Proper Extruder by Proper Printing?
Great testing as always. Now I get why the lgx (not lite) is a thing.
A lot of new printers come with those big gears now - QIDI's new ones use those, Creality K1 / K1Max uses those, and even Anker M5C has those big gears. But if I am not mistaking Bambu uses similar size gears that Orbiter V2 has.
@@DIYPERSPECTIVEdont forget the flying bears, those were the first ones with lgx clones, although every clone besides the k1 has spring loaded tensioners. The k1 only has one setting, lgx lite has two (rigid and tpu, tpu being the more closed) and the full size lgx 5 settings
I’m installing a 20:1 gear ratio ultra large gear extruder in my machine hoping to get enough torque to push 70-100mm3/s
I guess that is sarcasm? :D If not I wonder what hotend are you using, haha.
@@DIYPERSPECTIVE not sarcasm lol. Very very very long hotend. I can do 50 easy, but 70 starts to hit the limit of my Sherpa push force
The Goliath with VzHextrudort has done over 130 so 70 isn't particularly unbelievable. 🤷🍆
I wonder at what point you'd be better served by a geared BLDC motor & a rotary encoder for counting steps
@@daliasprints9798 with a 1.4mm nozzle yeah or with pla which needs less force than abs. With a .5 nozzle you need like 200n or more of force to push past 50 cubic
Haldis HGX looks like a nice extruder for the price
For one thing, big gears are pretty :)
The next logical step would be to test with TPU?
These videos take so long to make and TPU is almost a completely different topic. If the filament guide is close to the extruders' gears I see no reason why it won't perform better.
Pls try papilio extruder
Hmm, that is a very interesting extruder, so maybe, maybe. :)
Nicee i will wait before upgrade my only kp3s printer 🤣
something tells me that angled tooth profile is actually not good. sure you have more surface area, but less friction due to the angle. It could also cause twisting or the filament to move side to side on the teeth which would cause the same issues outlined by MirageC when looking at misaligned idlers. I've been saying this for a while now, but the next leap in speed printing is going to be switching back to 2.85mm filament. If we get a 2.85mm CHT Volcano nozzle with a 2.85mm ceramic heater Volcano block, and a 2.85mm LGX Lite or something similar, we will unlock the next milestone. Chinese companies: if you're reading, do it.
Ha... you fool, the biggest bottleneck is melting power, wider filament is worse because plastic has awful thermal conductivity.
Also: LGX Lite is weak AF.
PS: The Heater blocks aren't related to filament diameter, the nozzle and heatbreak are.
I got the vzhextrudort with angled tooth and one with angled tooth + helical gears. I havent had any twisting problems even with tpu.
Just a random thought, would it help if the angled teeth are opposite handed? so they are twisting the filament towards each other?
actually forget what i said, it's just going to rotate the filament in the opposite direction
Very good info but how do you explain the best extruder on the market the vz hextrudort… it’s small geared and proven to print faster than any other extruder in abs pla and tpu the softest filament there is. Your maxing out your hotend not your extruder. Do the same test on a fast printer with a high flow hotend and a cht nozzle, you’ll have a few other issues before your extruder slips filament in its teeth. And you have no way of knowing how much tension is on each extruder so if the smaller gear was a looser tension your whole study would be wrong I believe
You sound very biased for stating it is "the best". Link me the independent in depth testing results and then I believe you. I maxed out my tested hotend more or less at the same flowrate with all extruders. The only difference was that the big gears offered a more consistent extrusion overall. The tension was based on how much the teeth dig into the filament when rolled through them. Also are you sure vz extruder uses 8mm gears? Aren't they mid size like on the Orbiter v2?
VZ Hex is far from the objective best or fastest, see tests by Ivan Miranda and others, but you raise some good points in the second half of your post. Most of the high speed and high torque use different topologies that increase surface contact in different ways.
I wonder where the idea of angled teeth comes from ? 😅
Haha, but the thing is that you do not even need it on big filament drive gears :D COPY because others do lol.
Why are you so proud of that? It creates more problems and solves none.
@@unharmonix what problems?
After some more investigation, it would seem that the moter torgue is the limiting factor. Increasing the driver current and "tuning" the force spring setting, the performnace improved. The moter slips at time printing at 70 mm/s.
this all seems a little bit moot, and feels like an advert to sell a product by a feature that offers no real world benefit.
All the modern extruders were absolutely fine up to 28mm^3/s, if we're using a 0.4mm nozzle and 0.2mm layer height, that would mean we're topping out at 350mm/s print speed, except even the chc pro, with cht nozzle, and going up to 230c on pla can't manage 28mm^3/s on a 0.4mm nozzle.
And on print speed, if we look at the bambu x1c, doing "500mm/s" on all walls and infill and a reported "32mm^3/s", it will do an 8 gram benchy in 15 minutes, which actually works out at 9mm^3/s on average. it's only ever going to get near that 32mm^3/s value on long lines of infill.
"but what about bigger nozzles, they will need faster extrusion!", bigger nozzles need less force for greater flow rates, and we're already maxing our heat sink on the 0.4, so nope.
you also never explained how you reached the tension settings on the extruders you used, making the macro photos of the filament deformation absolutely meaningless; if you put more tension on a filament, it will deform more, but how necessary was the level of tension used for each?
In short, this "might" fix a problem when printers move even faster AND hotends melt even better than they currently do AND we either come up with a better way to melt plastic than cht, or we find a better material, as currently the plastic doesn't melt fast enough to go much further than we currently see, which is why super large format printers use a big pre-heated hopper, that gives time for the plastic to melt, rather than doing it all at the nozzle.
You understand that there are other big gear extruders on the market right? And nothing prevents you to buy other ones? Original LGX exists since the year 2021... And nowadays multiple new printers come with those big gear extruders like QIDI's, K1, M5C... My whole channel is based on genuine testing results, whether they are in favor of the manufacturers or not. I really don't care if anyone buys the items that I show because I am sponsored by PCBWay. And they just make all kinds of parts to allow me more easily and properly test things. It is literally a WIN-WIN-WIN scenario. If you think big gear extruders are not needed for high-speed/accel printers, use any extruder you want. EZ. :)
@DIYPERSPECTIVE I run a voron with a dragon high speed with cht hotend on an orbiter, and a Revo with a cht nozzle on a biqu h2. I've gone up to 2mm nozzle widths with 1mm layers, extrusion and flow rate are something that IS a factor in some of my work.
Heat sink rate of the plastic has always been an issue way before the extruder has started to struggle. I'd say I was in the top 1% of prosumer 3d print speed, if it's not an issue of viability for me, who is it an issue for?
Your video is titled that "small gears aren't viable". They 100% are. That doesn't mean larger gears don't have some theoretical benefits, like grip, but they DO also have theoretical drawbacks, like rotational distance accuracy and backlash; which are arguably more likely to show on a print than regular, predictable deformation before the filament gets melted.
And because you have a legitimate sponsor, does that preclude you from having an illegitimate sponsor too?
What really hinted that you have an agenda to push is when you compared the orbiter, which was within the same margin of error as the extruder you are reviewing, yet you implied it was somehow not performing as well. Probably why you left those results up for only a couple of seconds 😂
@timreeves do you not see the question mark? ?????
@@coltenmeredith8899 did you not watch the video where he concluded that small gear extruders were in some way a bottleneck to a 3d printer? Which they're not.
@@timreeves You mentioned the title. I said you mentioned the title. Yes, I watched the video
Dual gear extruders suck because you can´t have an adjusted gear and mesh the teeth perfectly.
Single gear or decoupled gears which are large or a belt extruder are best. And why use gears at all ? It wpuld make sense to create spikes instead to drive the filament even better. We use gears because that´s what out there on the marked.
I see you too watch MirageC videos lol. The only option that I know without gears is the Proper Extruder. And the one that could be good with one filament pushing gear is the Nextruder if the Prusa ever sells it as a standalone.
@@DIYPERSPECTIVE I mean it is clear even without his videos but yes he has shown that I believe. It´s technically impossible to use dual gear for the filament AND have those both gears mesh. This idea only works because of the slop in the system and it is fine but if you want super precise extrusion, it just does not make sense.
The best extrusion possible would be archieved with a worm gear, just like in a injection molding machine. Which would also replace filament, it is generally a good idea since filament is about 10x the price of standard plastic. But it would be also difficult, expensive and heavy.
Second best option is to use one large drinving gear with gear reduction and the mating surface to the filament should have really sharp prongs instad of the normal gear geometry. Those would pierce the filament and push it. I think there is one company which already uses a similar concept. Even saw a wood mill ? This is also how they transport logs.
Do you test the theories in reality or do you just speak "in theory" stuff? I am genuinely interested. Because most comments from you are so negative no matter the topic. It is like the industry is doing everything wrong and you are the mastermind of some sort. Sometimes you make interesting takes but also some of them are complete garbage like recommending new QIDI printers, which are plastic trash boxes with an integrated CoreXY system. So it is hard to take things you say more seriously, you know.
@@DIYPERSPECTIVE As a mechanical engineer and developer I might see things differently, that´s why I point out negative aspects. 90% of all comments are negative, what´s the point in writing "there is no problem whatsoever" How do you improve this way ? There is no value in patting yourself on the back constantly.
And of course not everything is bad in the consumer 3D printing market, far from it. The machines are incredible cheap and power. There are many machines and even different technologies like FDM and SLA (or DLP). The price to performance is fantastic with almost any printer these days.
And while most machines are of great value and capability, many manufacturers (especially from China) make such stupid and unnecessary mistakes, it´s hard to believe they didn´t simply copy + paste a design. Same goes even for Prusa.
I have been saying form their MK1 on to change the stupid threded rods and stupid stuff like this, and it took them years for this simple modification.
Also you are right, recommending the Qidi printer wasn´t great on my part, however it was not my intention to recommend this specific printer I just wanted to show that there are other better value options. I am not a fan of any brand I only compare capability and price that´s all.
RIDGA provides quality that is almost perfect. I see no reason for a derogatory comment just for the sake of doing it
I'm deeply disappointed in mine. It slips constantly. I'm printing at 70 mm/s. I find better performance with very little pressure on the springed lever. Only just screwed in. I'm going back to my TitanX. It could be that I have a "dud" unit. I know from building my own, the feeding gears have the be centred perfectly around the filament hole.
Who cares !! High speed printing at high flow rates.
Poor Quality prints .. generated at faster speed in some twits' basement/shed . ..
That's Some major techno revolution..🤣
High-speed prints don't always have to be bad quality.
That's just foolish. If your extruder is "forcing" filament through the nozzle, then it's coming out only half-softened, and that's a recipe for a really bad print. Why can I get very high volume through my single-small-gear extruders when you cannot? There's no good excuse for double-gear extruders and there's certainly no excuse for large-gear extruders. They buy you NOTHING.
even if your plastic is perferctly melted an increase in flow through a constant orifice will lead to higher pressure, thus a higher force required to extrude the material (furthermore this relation is not linear but squared, so it does make sense to optimize the extruder along with the high speed capabilities of modern 3d printers)
@@orphax1925 That's silly. I know people who run race printers at ridiculous speeds. Their extruders never need to force filament out the nozzle because they run appropriate temperatures. There are lots of people out there running 500mm/sec with a single-small-gear extruder and getting excellent quality prints.
Maybe if you run at 1300mm/sec you might need a little "forcing"... but if you force it, then your print quality will drop off very rapidly because your filament will spray out of the nozzle instead of being extruded cleanly in a well controlled fashion.
Don't just take my word for it, though. Put a single-small-gear extruder on one of your printers and give it another try, giving it your best shot to remain unbiased. Don't just promote expensive waste because you can. These aren't Barbie dolls, they're 3D printers.
to add to your point I'm not convinced that biting into the filament harder is a big deal. sure it gets longer as you bite harder and deform the filament more, but it gets longer after it's already dosed. just crank down the tension adjustment knob
I don't entirely agree that large gears are pointless, but I think they probably don't really shine until you're running a supervolcano type setup.
@user-hy7cg9jg5r that's the most ridiculous thing I've ever heard lol, for a start "race printers" like Vorons, Ratrigs etc. all use dual gear direct drive extruders to begin with, maybe in your mind filament melts and just falls out at 500mm/s but in reality it takes a good amount of driving force which you'd know if you'd ever actually printed at that speed instead of just regurgitating things that other people said. As for your extruder remark if you ever intend on printing more than just PLA trinkets and knick knacks dual gear is an absolute benefit, try printing SBS with your single gear plastic extruder, you won't even get past 50mm/s nevermind 500.
Just because you're unwilling to spend the time/money or unable to see the benefit doesn't mean it's not better, it just means you'll never progress, while everyone else has fun with current gen stuff, you're the guy that goes around telling everyone his stock Ender 3 prints faster than our coreXY printers because you're just so clever.
@@JoshW1ck You've apparently never seen an Ender-style printer running happily at 500mm/sec. My condolences to you; your experience is lacking very badly. Maybe after you actually see it happening you'll have a different perspective.
SBS filament? Again, my condolences. Nearly everything I print is in either PETG or ASA (I print working hardware and pistol frames), with a little TPU here and there. Why in the world would you buy SBS? Did someone convince you to just stand your wallet open on the street corner?