ฝัง
- เผยแพร่เมื่อ 28 ก.พ. 2021
- As I am super excited with this project, I could not keep it entirely to myself....
So here a sneak peek to my next official video :)
Stay tuned!
NOVA Hotend ► 3dpassion.com/nova
You want to know more about the HevORT advanced DIY 3D printer?
WEBSITE ► www.hevort.com
You want to learn and contribute to the project?
FORUMS ► forums.hevort.com/index.php
FACEBOOK ► / hevort
DISCORD ► / discord
🍺 SUPPORT ME ► www.paypal.com/donate?hosted_...
If you like my work and would like to support my efforts in making more open source 3D printing solutions, please consider a small donation. Thank you.
MirageC - วิทยาศาสตร์และเทคโนโลยี
always impressed by the speed holy moly haha
Wait...you mean to say that that wasn't sped up? That was real time printing speed? Holy moly indeed, wow
@@BinaryBlueBull it's speed up video. You can hear from servo sound
@@GameBacardi That's what I was thinking initially as well. I assumed by reading the comment we're replying to that Chuxpie saw this printer before and knew it was the real speed. That's kind of reassuring, I was proud of myself when I managed to reach 150mm/s with my printer while still printing perfect quality, and then I saw this video and thought "damn, I'm proud of 150mms/s and this guy is going warp 9"
@@BinaryBlueBull the original uploader himself says its realtime and the video doesn't look sped up to me so I doubt it's sped up
@@MaxC_1 Indeed I read his reply as well. I'm completely new to this channel so I had (have) no idea whether or not to believe that. Now I don't know what is true anymore, haha. It would be absolutely insane if that actually is real-time speed, extremely impressive in that case. It would also be a huge compliment to the person that built it, since there's discussion in the comments about whether or not the video is sped up. I did see this kind of speed before once or twice so I know it isn't impossible at least but those were bigger prints with much longer straight lines and very gentle corners, the above print is much harder because of the sharp corners and the small size of the print
Yea impressive hardware design 👍
Thanks for sharing your experience with all of us 👍😊
That looks brilliant. When is the long video coming on this? I can not stand such a short teaser ...
What I really wanna see now, is leveraging the 3 axis of the bed for true non-planar printing. Software wise, maybe something like "spherical slicing" would be relatively easy to implement, and that could be coupled with the non-planar technique we've seen developped at University of Hamburg a year ago.
this hotend would be awful for that
@@fish9468 for the Hamburg nonplanar technique, yes, but not for the technique I'm proposing. Tilting the build plate as you print would mean the nozzle stays flat / tangential to the spherical slice. You'd be printing kinda like the top of an onion, a so called "spherical cap".
Was just looking at non planar slicing to see if there is anything uptodate and consumer ready and still disappointingly few advancements, found Curvislicer though which is the closest one after the Slic3r method, since i found a hotend i hadn't seen before from Protoworx, hoping non planar comes to normal slicers at some point without having to build stuff from sourcecode
@@Tarex_ There's an addon for Blender by Alessandro Zomparelli called gcode-exporter. It's not a slicer, but can convert edges or polylines to gcode. If you're comfortable with Blender and gcode editing then you can make something that sort of works.
@@Anthromod i use Blender and Gcode is fine as long as i don't have to write it by hand :D but awesome, will go get the Addon right away, the new nozzle arrives tomorrow, will see how i get Blender to slice a surface into polylines, guess Edge tools, Thanks for the Tip!
This looks amazing for not only keeping excess weight at a minimum but also keep the center of mass close to the rail for better high acceleration handling, very cool.
A new modern twist of a Wades extruder, looking forward to seeing more.
There aren't many other effective ways to push filament other than a rotating gear (or two) that pinches the filament to push it through. What's new here is the merging with the X carriage to save some weight and have a better CG management.
@@MirageC I guess another way to push filament would be a self-tapping internally-threaded bore that could be rotated, but then that would likely cause twisting of the filament.
@@MirageC 'aren't many other effective ways to push filament ' The one other effective system I've seen that has an advantage or two is the belt drive extruders, that use the back side of timing belts as a friction drive. More grip, doesn't damage the filament surface, etc. But it's hard to argue with the ease and simplicity of dual gear drive and it's fine for most filaments.
@eldritch check fuselab 3d they made exactly that.
@@ModelLights And now I'm picturing a hybrid of the two designs, which is essential a baseball pitching machine as imagined by someone who designs miniatures for model train sets.
To get more flow rate per weight i think a BLDC motor coupled with a "harmonic drive" type gear is the way to go. High speed BLDC motors produce the most power per weight and with a strain wave gear you can get a really high reduction in a small light package.
I would rather focus on precision and cost reduction and affordability. speed is already achieved.
Do not focus entirely on one aspect when you trying to create "best 3d printer"
You don't need a lot of force to push a liquid, do you? :) But you do need that force to push a semi-liquid :) The answer is in details.
weight plays a crucial role in deceleration, printing speed, and print quality
it would be helpful if you could share the weight comparison of a typical direct drive and if possible, a comparison of optimal printing speeds
powerful, I love this design
nice, looks similar to the sherpa mini however you seem to have further integrated into the carriage. Ultimately the question then arises how to force feed the weaker dde with assistance from a 2nd bowden extruder. And keep the feed / retractions in sync with each other etc. Which is the next logical evolutionary step in order to support higher flow rates. BTW on that same subject... will you eventually be wanting to try the new slice mosquito magnum+ high flow? And comparing it since you already have the nova hotend. That would be an interesting competition to see who wins, under a fair and controlled conditions.. ofc slice have not started selling them just yet. So that is still 'some weeks off' until a widespread availability.
beautiful!
How long does it take the filament to reach the desired temperature? At high speeds wouldn't the filament be too cool from being pushed out too fast for the hotend to properly heat it up?
how is the finish and tolerances on the prints ? Impressive speed
The music startled me a bit. I'thought just for a split second that you extruder grinds the filament;). Love your work!!!
Subscribed! Very interesting video. Is that a timelapse or realtime? 😲
Real time of course! :) thanks for the sub!
So uhh, when is the video coming out on this exactly? I'm looking at an extruder upgrade very soon here!
Could you provide the priming gcode? Looks nice with the little pile at the end :)
Ah, the Binford Tools method of machine improvement.
Well, I had a talk with Wilson and he is kind of okay with the idea ;)
Hello are there plans to do a mount for a mosquito hotend or are there reasons not to? Cause i will build a HevORT this summer and wanted to go with a direct drive Extruder to print Materials like TPU or Nylon. Ive been already thinking about doing a remix with a sherpa mini and a Nema 14 stepper.
All details now available at: miragec79.github.io/HextrudORT/
I am torn with your idea of using the filament cooling not on the extruder, but on the bed. You already made a video on the topic, but what would happen, if you use an encloser with no outside air, just circulating it inside of the chamber?
The same as with a on head cooling system into an enclosure.
I don't know what I'm looking at, is that a round stepper motor? Is this setup a product for sale?
Extruder itself is not his idea. Search 'Orbiter extruder' at thingiverse. As far as I know it was the first extruder to use this motor.
Then search 'Sherpa mini' on github. If i'm not mistaken it is based on Orbiter, but simpler and easier to get parts for it. ( it uses BMG internals)
What extruder do you recommend to use with volcanoes hot end?
logical next step I am woirking on is to move away from steppers toward dc or brushless closed loop servos. they can be really light small and powerfull, loosing steps gets impossible and you could even get force feedback!
@Flip Schwipp, me too. What driver chips are you playing with?
Yes, servos are the next big thing in 3D printing I my opinion.
@@MirageC I misread the original comment. I thought @Flip Schwipp was talking about extruders, not printers in general. I'm working on a bldc extruder atm to replace steppers, mainly in an attempt to minimise mass at the print head.
@@simonmerrett Well I think all steppers could eventually be replaced by servos or brushless motors. Keep me posted on your progress, that sounds interesting.
@Kenny Eaton I'd consider trying with Novec 1230 as the "air", because it has far higher (thermal) mass than nitrogen (per volume). It's boiling point is 49 C, so unless you squirt liquid on freshly-extruded parts, you'd have to run it at elevated temperature.
In general, servos will get you far more acceleration, so fine details will have a lower speed penalty. This requires a very well balanced gantry, so the acceleration along the X axis won't twist the head around the Y axis (and vice-versa). Because linear deflections due to acceleration are easy to predict and compensate in software, but twisting causes non-linear interference between the axes. One has to be careful so the drive signal doesn't contain much energy near the resonance frequencies of the gantry, but this just shifts computational effort to the conversion from G-code to concrete motor ramps. Your Arduino won't be enough, but something a bit more powerful like a RaspberryPi-class device should be enough.
And even that is just to deliberately attenuate movement close to resonant frequencies without getting janky. But once you have a well-balanced gantry, resonance-suppression, and compensation for deflection-due-to-acceleration, you can use very powerful servos and retain accuracy at over 100 m/s^2 of acceleration.
You need to be light (which means things will bend), or each servo needs it's own breaker due to the powers involved.
Can it use a 1mm nozzle for insanely fast prints
Cool! I'm trying the remote direct driver extruder with flexible shaft. A think it's a better option.
I did that. And was not happy with the results. Mind you it was a V1 Nimble. I hope that technilogy develops and that the rippling artifact goes away. It is a very good idea and has a lot of potential.
That's pretty cool! I'm still confused why no-one has tried a screw-drive system with three worm gears in parallel with the filament. That would give maximum traction, especially since you can just make the worm gears longer if you needed more.
An odd number of gears, if they are meant to mesh, won't. Am I correct in thinking only the filament would touch the worms as it spins? The problem I could see happening if I'm imagining this correctly: 3 worm gears with a vertical axis of rotation for each, in a triangular arrangement, with filament going down the middle, is that the filament would have spin imparted on it, and it'd turn into a giant twisted knot very readily.
Another poster mentioned instead using timing belts, which would make more sense. Another option I can think of would be using 3 loops of chain instead, but not sure how much backlash you'd have if you did that.
Now you've got me thinking, and that's when I start causing trouble :D
@@strictnonconformist7369 Hmmm... I hadn't thought about turning the filament. It would need to be an even number, then, with an equal number going in each direction.
I thought someone had tried something like this a while back?, maybe an old RepRap forum project, I could be wrong though.
Replace the flex cable with a square shaft stretching the width of the gantry tied directly to the extruder stepper mounted next to the x axis stepper and a bearing support on the other end. The extruder directly mounted over the hotend with a sliding square drive input gear. Minimum weight to the hot end, maximum speed, less flex from the flopping flex cable. The trouble is finding the right material for the sliding gear and the driveshaft, but it will give you the best of both worlds if you do.
No Flex cable involved in here. Stepper motor is on the print head. And this is a CoreXY machine. The weight of the stepper would need to be carried along with the Y axis.
@@MirageC Would still work, it would reduce the weight of the stepper being flung left and right, although I do admit I have been dreaming this up on more of a low end Cartesian style printer, but this is the next logical step along the line of a remote direct drive :) ANd I mistook that for a demote direct drive my apologies!
Looks familiar. Does it have something in common with the Orbiter extruder?
Voron had similar setup with similar capabilities available for a while now. With Orbiter extruder and high flow Dragon or Mosquito hotend. Still, it's nice to see how different people solving the same problems can arrive to similar conclusions.
Yes it shares the same LDO motor as the Orbiter and the Sherpa from Annex Engineering (designer of K2 CroXY printer).
To me this looks functionally the same as a dual geared BMG, unless it is a smaller scale (hard to tell from video), just tightly integrated into the carriage itself. Is that kind of the idea or is there more to the story?
That is pretty much it yes.
oof! nice! what material are you printing? no part cooling at that speed?
wait, is that the new extruder you designed already in action? woooah... printing is realtime or speed up in video?
Real time of course :) yes its the little guy I was working on lately.
This PLA. Part cooling is achieved by my HevACS system: th-cam.com/video/65FVQ1jArME/w-d-xo.html
Can you share the info (Current, acceleration, jerk and speed) from the Extruder motor you are using?
And, does the motor get too hot?
Tnx!
Of course! On Duet RepRap Firmware, I used:
E Accel: 2000
E Jerk: 1000
Max E Speed: 80mm/s
motor current: 0.350A
On Klipper I use the same current value. Motor is getting medium hot, I would say in the 40C while printing. I do have side duct cooling helping a bit though, but these little motors are rated Calss A (180C) so they can generate quite a good amount of heat. Make sure you install them on a heat resistant carriage.
@@MirageC I bought a orbiter extruder, and forgot to change the current... Made a print test, but had to go away from the printer. When I returned, the motor melted the gear housing...
I hope the motor is still good. But unfortunately the gears are toasted! Hehe
Live and learn... Thanks for you info. It will be useful!
@@tiagopviana sorry to hear that :( I almost did the same, those motors are hot in all ways! ;)
@@tiagopviana ouch!
this does make me wonder how strong the direct drive extrusion still is at this kind of speed especially with that total 15:1 gearing. At low speeds of course very high but at this speed i doubt it as steppers lose a lot of torque above a certain step frequency. maybe the limit is much higher than i thought
Not sure I am understanding how you get that 15:1 gearing ratio? In the end this is a 1.8deg motor with a 5:1 gearing ratio using BMG gears. It takes 707.35 steps/mm @ x16 micro stepping. BMG is 415 steps/mm for reference and uses an approximate 3:1 ratio.
@@MirageC oh, I see, thought it was a 5:1 stepper first and another 3:1 gear after it, my bad. Still am wondering if it impact torque significantly. I thought most steppers start losing torque above 500-600 RPM. I could be wrong though, just my 2 cents, seems like a fantastic lightweight design!
How is this different than any other direct drive?
No time, to print slow :-))
Great :-)
what is your 3D modeling program?
Fusion360.
How is this with really flexible TPU?
No clue yet. Will have to try.
will there be a kit to buy?
Yes! working on it ;)
@@MirageC Nice! Thanks :)
Interesting!
I'd love to see a PCB and E3D toolchanger based extruder. Instead of milling aluminium you could order thick PCBs for the toolchanger plates and engineer the direct extruder to be build directly on top of the PCB. I'd also like to see DC motor and encoder driven extruder for less weight. I don't don't if this could really work of course.
Wow! this is very good Idea! PCB material is super strong, and perhaps we could use it for signal and power distribution! Thank you for the idea!
@@MirageC Thanks! I wrote some more about my idea on the E3D forum. Integrating electronics could definitely be interesting too. Like short runs to the thermistor and heat block. Or even integrating an MCU too and only having power and signal wires.
forum.e3d-online.com/threads/pcb-based-automatic-tool-changer-idea.3966
I actually tried this a few years ago. FR4 is not a good choice for tool changer plates. It flexes as hell and is too anisotropic for this application.
@@ericvauwee4923 Oh interesting. I would think that something like 3.2mm would be thick enough for 3D printing application, but of course it's really not made for this purpose
@@dejayrezme8617 Yes you propably get away with 3.2mm for 3dp applications. As I wanted to implement a "universal" solution to access also subtractive and other tools that require an elevated rigidity, I abandoned the PCB option after the first prototype. if you solely focus on FDM tho this seems a good option.
will like to see if it's possible to make the puny 28byj usable on this hotendsetup
I think these would'nt be able to provide the speed required. especially for retraction,
I guess its using a brushless motor. That should be the only way to go this fast without losing accuracy.
Stepper Motor Nema 14 with a gearing ratio of 5 to 1.
the acceleration on this one is like over 9000G
9000Gs would be a bit extreme ;) (1G = 9810mm/s).
This print was using 18000mm/s^2 acceleration which translates to 2Gs only ;). With Square corner set to 5.
@@MirageC whew. only 2Gs. it could have been much slower.
Looks mega! Could you make me an adapter for my Ender6 for it? I also have a Hextrudort extruder and also want to buy a Nova hotend :)
Looks like sherpa mini
I am using the BMG internals and the LDO round Nema 14 as well.
have you considered closed loop BLDC?
No, not yet. I am staying as close as possible to motors that will be easily driven by conventional 3D printer control boards due to time constraint, but I am open to any suggestion. Do you have any particular BLDC and controller in mind?
I am doing some research on this since a couple of weeks, but I lack the time to do something meaningful tbh. It’s very interesting, since bldc motors have a very high power to weight ratio (technically speaking nema stepper motors are bldc motors as well, but I am talking about high powered rc motors...)
I found a motor that should have enough power for an extruder, while weighing 15 grams!
@@MirageC Not a particular. Because of better magnets BLDC give optimized weight/power ratio. I know that people use Odrive controller with BLDC motors with axis when they want tons of power and speed. As comparasion nema 17 weight about 200g and can handle about 5W continous power. At RC bldc motors you can get from 40g motor about 400W power. So power to weight ratio is something like 300x better. Naturally that is not so simple thing when counting that BLDC will need buch bigger gear ratio (most probably planetary or harmonic) but I think it still can compete with normal stepper. I dont know is there official step motors what have better magnets. As Max say at thechinally speaking bldc and stepper is same thing.
Are you using Klipper to achieve this speeds?
Yes
wtf 0_0 that nuts
Hello. Do you have an encoder on your printer?
You mean rotary encoder for stepper motors? No i don't have.
@@MirageC Yes. It can sometimes skip the stepper motor step and the layer can be separated.
@@martinTintin476 He has some insanely strong stepper motors instead.
@@traniel123456789 Attention, but stepper motors! Every printer has this, but few that have insured skipping a step.
I was thinking if I used the 2x speed to watch the 3D printing.
How do you like the Nova so far?
I simply love it! Because of its tight tolerance around the filament it reacts more like a syringe than other high flow hot end I have tried. Its super small, super light (45g all loaded with fan) and powerful. Can't ask for more. I thing it is a great balance between weight and melting power.
@0:13 sounds like sonic is about to drown.
Why it is that fast?
the speed wtf
Did you 3d print while extruder....?
If yes what your settings... 🧐
Yes i did print that extruder. 0.16 layer height with 4 walls and 15% infill.
@@MirageC now im using delta printer with 220mm × 350mm build volume
im thinking to build coreXY printer with 2020 aluminum profile if i print all joints with PLA will hold the printer at higher speeds or should i go with aluminium corner joints
Those joints higher price 10pcs of corner joints like 5$ i want total of 35 parts im thinking print those will hold...!
Holy f***....
My hand doesn't want to press like. I like this device, and I don't like it at the same time. Sherpa is elegant, but still far from perfect. As for the heater used, I just hate it.
May I ask what went wrong with your Nova?
I have been checking out the Nova, what makes you hate it?
@@gloopann the price... lol
@@yuripolak compared to what?
@@MirageC regular cheap clones... I know the performance can't be compared. I even think this should worth every penny, just not for my budget yet
Onshape?
Fusion360
isnt it just a normal direct drive?
yes, but instead of having it as a separate unit that you have to attach on with fasteners, it is embedded into the carriage. Better center of gravity management and some weight saving. Not super revolutionary, but definitely worth it when comparing to other brick shaped extruder available on the market now.
Can I ask why the Nova? I haven't seen very much info about it. I know it exists but it's kinda standing in the shadow of a Mosquito.
Had the same thought as well.
Nova can flow 50mm^3+ and it weighs 45g, I like it very much.
@@MirageC As a Nova user I can support these claims
Agreed, the Nova makes some ridiculous claims, I would like to see a fair comparison between both.
@@certified-forklifter not so ridiculous when you have it on your printer :) I have recently done 62mm³/s with it and shot that on video.
I'm impressed that it's running so fast. Maybe you can improve consistency in quality with linear rails?
lol, now I'm dizzy
You make us watch an ad for this?
@@wojtek-33, nope.
Creators decide to show us ads, not TH-cam.
@@wojtek-33 well, now you know
Super! Ultra szybkie tempo :) pracuje nad tym w mojej drukarce aby osiągnąć takie rezultaty :)
Where is stepper motor?
Where is the part cooling
@@maxmustermann2596 check another video
It is using NEMA14 stepper (smol round one) with, like, 7:1 reduction.
@@LetoDeWirre nice
@@maxmustermann2596 look at my video about part cooling. I use side ducting with 120mm fans: th-cam.com/video/65FVQ1jArME/w-d-xo.html
Hello Cousin :) Comme d'habitude, une démonstration like "HellvORT" ;) Thank you from FR
Merci à toi 👍
I have so many questions...
but mainly: wtf is going on?
Simply trying to cut down prototyping time for the hobbyists and the small businesses at a faire cost :)
someone gave an extruder a mix of meth, cocaine and pure caffeine. google 14 minute benchy.
@@SignalJones @Nitram with is delta is chasing sub 5minutes! 5:45 minutes in this video: th-cam.com/video/J8p46wDM5KM/w-d-xo.html
my ender 3 v2 did not like it
:)
Note the unnatural camera shake for 24-26 seconds. The video has been sped up. And the sound of motors at such accelerations will be different. Draw your own conclusions.
Yes! And the earth is flat....
did I accidentally watched it on 10x?
If you did, you would have missed it in the blink of an eye ;)
I was hoping it had a flexible driveshaft like a weed wacker when I clicked on the video. Direct drives have too much inertia.
I did try flex cable before. And could not get away from the ripple issue cause the the winding and unwinding of the cable under load. That would be a killer setup, but unfortunately never seen one working great :(
@@MirageC Glad you gave that a try. The person who makes such an idea work is a hero...
Looks like you borrowed a few ideas from the orbiter... Also, too fast paced. Couldn't properly look at any of the animations or drawings because they flashed by too fast.
Sneak peek.... more to come with A LOT more details :)
so we are just going to ignore the terrible quality?
Quality of what? That first test print?
Blank page @t
This looks sped up.