@@kendell8046 Wouldn't that actually result in a slower print speed? I don't know if you would be able to eject enough material through a small nozzle fast enough to get up to these speeds. Also, I think you would need a bit of a higher temperature to get the material to flow well enough through a smaller opening and not clog. But I could be wrong here, just a guess.
Due to the short line segments in such a detailed model as benchy, I don't think it will accelerate all the way to the topspeed, so it wouldn't be a comparible result
There's not much difference in a model like a benchy, as the printer won't reach max speed on any segment. That's why the model he's used is a bunch of long straight lines, to give the printer a chance to accelerate to the target speed. It might be interesting to see a giant benchy though, scaled to the biggest side that would fit on the bed. I don't think it would be able to reach max speed but it should reach higher speeds than a 1:1 benchy.
@@suharsh96 Well then, if we're not going for quality here I'm sure I could crank out a benchy at 400 mm/s on my printer. Hell, i'll just tell it to shart out a bunch of filament in a pile at 800 mm/s and call it a benchy speedrun.
The thing is that these speeds are not accurate in the same way cpu clock speeds are read. There is no way to measure the speeds, sure anyone can order their printer to print at 1000 or 10,000 mm/s but the acceleration is still too low to achieve it. It certainly is moving very fast, just not 700mm/s fast.
In no way can you compare this to liquid nitrogen cooling! Because speeding ya printer up is actually a good idea To be clear, this comment is a bit of a joke.
I'd love to see this beast printing a simple cylinder in vase mode. Should be able to reach even higher speeds without the cornering issue. Great work.
Yeah I was thinking the same! Would be really cool. But I guess the lesson for this video is that you need to work on the melting in the extruder. It would make more sense to print at 400m/s with 0.2 layer height than 800m/s at 0.1.
@sourand jaded and it's not like the control software is going to give a damn about the angle of the bend. infact it doesn't even know. make a circle. slice it. and you get several thousand lines of code for tiny individual lines. (PER LAYER) without ANY info on the angle of the next line. and it's just coördinates. G1 X0 Y0 G1 X0.01 Y0.0 G1 X0.02 Y0.01 G1 X0.03 Y0.01 G1 X0.04 Y0.02 excetra excetra excetra and it is going to treat them as such. you can tell it to go 800m/s all you want. it won't go faster then 50m/s (depending on the resolution of the circle) you can then change play with the speed setting as much as you want. it's going to print it at the same speed each and every time. however, if you change the jerk and acceleration, THEN suddenly it starts to go faster (or slower if you lower them) if you have a 32bit motherboard and some special code in the firmware. it could look ahead in the g-code and begin doing some math. (basicly turning all those G1 straight lines, into a G3 or G4 circle, then it suddenly does have the info of the angle on the corners, and it can decide to go faster because it doesn't need to slow down as much) you can also install a plugin into the slicer. and as long as your firmware on the printer knows what G3 and G4 is. the slicer can just write (G3 with the coördinates of the centre point of the circle. the coördinates of the end point of the circle. and perhaps a radius) and it would be faster aswel. and then you can print faster circles on an 8bit motherboard aswel. next time you don't know jack shit about a subject. try to just ask a question. rather then pretend you know something and make a fool of yourself. or just shut up. either or. learn how to program a lathe or mill with fanuc iso control before you come back. (not heidenhein. that's the easy mode) marlin and most 3d printer firmware is a childs toy compared to real cnc. and 3dprinting shortcommings show this well.
This just shows that 3d printing can progress to a speed that would print in acceptable times as the technology improves. We are a long way off high speed prints for a while though - 3d printing is still pretty much in its infancy at this point.. I can only imagine the stress that all the motors are going through, this has to impact on their working life.. and the noise is something else entirely!! Who knows how far we can go in the future? Who would have guessed 50 years ago that we could have computers the size of a matchbox and 3d printers making all kinds of objects, some even for medical applications? Yeah, stuff of science fiction, but we will catch up.. Good video, well done..
Imagine in like 50-100 years when we're all printing live tissue with this sort of speed. Want a cat? 5 minutes, tops. Accidentally lobbed off a few fingers or your arm while working on the ol' hover-mobile? Go to Fiverr and have someone make you a new one.
Anyone remember the first 2d printers? These things were slow as heck, jammed all the time and required constant supervision - sounds familiar? Probably we'll never achieve StarTrek replicator speed of 3d printing (because of that pesky thing called basic physics) but I am sure that we'll eventually achieve 3d prints that are done in minutes instead of hours and have acceptable quality. Maybe even full color 3d prints where printer would just melt together CMYKW (W=white since it cannot rely on the medium such as paper to supply that color) filaments in right proportions to get proper color.
@@UltimatePerfection in the movie "face off" (John Travolta/Nicolas Cage), there was an ear being 3d printed in a fashion very much like resin printers. That film was 20years ago - who would have imagined that 3d ears and hearts are being made in today's medical industry? You are right about 2d printers and the resemblance.. Who knows where well be in future years..
Man,bless you for sharing your experience with the rest of the mortals. You should be the President of the 3D printing industry for pushing the limits beyond limits. Thank you and salutations from France.
yeah, we need more people like this! what are we, like 5 or 6 people? LOL this needs to take off, I got tired of waiting for prints, and I'm printing at 150 🤣
This is impressive. I’d love to see how it does laying down a complete bottom layer and maybe a more complicated part with some retraction. Glad you were recommended to me today. New sub.
I wonder what the theoretical limit on speed is. At a certain point the filament would have to be near-liquid just to come out of the nozzle fast enough, and eventually it would just be a jet of molten filament coming out like a water jet knocking over anything already printed. At that point you might as well just stick a mold under it and call it an injection molder.
@@DUIofPhysics I'm melting more than 200g/h with my Nova hotend (47mm³/s or so) and I can tell you, neither specific filament or a ridículous long heatzone are the way to go, I can push that stupid amount of (relatively good) regular plástic (I think it's the same ingeo resin as prusament) and the Nova has a heatzone of about 18mm, sooo
Dude, my hats off to you. I've got almost the same hardware as you but haven't added the volcano yet. I was stymied at 100 mmps by the melt rate. If that is what I can milk outta a stock volcano, then I can't wait to hotrod it. 🤯
This is completely insane. I am glad that you are one of the only persons that actually try and show what you can and what you can't do. This one of the most annoying parts of many DIY machine communities to me: Nobody wants to test or talk or show about the qualities of their machines. People literally say on their communitie's subreddit about their own machines "it's a 1000+ dollar machine, of course it prints very good". It is really, really frustrating.
@@MirageC :D (Just to clarify: I meant that people are telling that in their subreddits about their own machines - I edited my post to clarify that ^^. That's why I think it's great that you are testing and showing the limits and results from your machine here. I remember asking you in a previous video about that. You said you are planning on doing so. And you did it. And that's awesome.) - Your link for us viewers: ruclips.net/video/aL7pEEHTTe4/видео.html
no-trick-pony_lockpicking thank you for the honest word. That’s how I like to think. Nothing to hide here ;) just trying to move the 3D printing world one notch forward ( or confirm dead end! Hahah)
Incredible mate 😍😍😍 HevOrt is one of my favorite projects to follow and you never ceases to amaze with design, iterations, changes, and test videos that are absolutely outstanding !!! Incredible work 💪💪💪 !!
First time seeing this project. As an engineer I was already thinking from the title "how the heck could the tension belts and motors handle that speed". Your design solves that problem. I'd imagine you'd run into severe stringing issues if you were printing something that required travel. Following👍
That is a very good test idea. Stringing at that speed. Never had an issue with stringing though. Pressure advance does most othe work. A slower retraction also help in order to avoid the fusion bath to break and let some material fall loose from the hot end.
@DocH same thing that happens with computers: “this much RAM is so much, I’ll never use that much! This much hard drive, more than I’ll ever need! This fast of storage, faster than I’ll ever need (now moved to SSD)! This fast of CPU, faster than I’ll know how to use! This many cores, I won’t have a way to use all that!” Performance inflation, same thing as lifestyle inflation: too much is never enough!
Cool project-I love how you built your printer, very close to how I want to eventually build my own at some point. The speeds you reach are incredible! You got a new subscriber. ^^
Great video. I like the way you new our to correct the problem each time. I have printers but not the high end like you got but I have learned something from it. Thanks and great technicians work.
We need more people working on this along with Top Down Resin printers... Right now my DLP printer is going through testing, at 10 cm per hour on the Z height with a build plate of 384 x 216 mm. Hoping to replace the LED in the projector to cut the time down to 35-45 minutes, the only problem then will be resin viscosity and wiper splashes.
wow. incredible. you're even lugging around that heavy hemera and x-axis rail. I feel uncomfortable when my super-light bowden delta goes above 250. new sub.
you would probbably shave about 1 hour off of that time with speeds like this. you can tell your printer to print at the speed of sound. but if it doesn't have enoegh lenght in a straight line to ramp up to said speed. it will never get there. i can tell my printer to print at 800mm/s but then print a cilinder. and it will only print that thing at about 100 maybe 150mm/s why? because jerk and acceleration are limiting it's speed. because that circle is hundreds and hundreds of tiny straight lines that it keeps needing to slow down and accelerate for.
@@darkracer1252 I think that's a great point. I have to ask, as I'm somewhat noobish, wouldn't the absolute maintainable speed (I want good print quality too) be directly dependent on the geometry of the part you are printing?
@@TheJacklwilliams yes it is. and as cris mawson says. arc welder helps (it's a plugin on cura that turns circles in circle commands rather then straight lines) so the thousand or so lines of code for the short lines that make up a cirlce will be turned into a single code. but what cris didn't understand is that that doesn't change what i said. jerk and acceleration has nothing to do with the amount of code. arc welder only helps if you have an 8bit controller. it's STILL thousands of tiny straight lines that all require it to take jerk and acceleration into account. if you have "junction deviation" set up though it will slow down minimal because it takes the angle of the corner into the equasion. so that's a jab at you cris mawson. arc welder is solution to a slow cpu. not to slow jerk and acceleration.
The tablet is displaying the native Web Server from the DuetWifi board I am using. This control board will connect to your home network. You can then access it using your smart TV if you want :)
I can't figure out how your printer is so quiet at those speeds. Is tmc stealthchop working for you at these speeds? I (and many others) am battling with stealthchop being very noisy, to the point that spreadcycle mode is more quiet at >100mm/s. Which motors and pulleys are you using?
I want to see how extruder motor is spinning at those speeds 🤯. I would be perfectly happy if I could reach 150, at 400 I was blown at 600 my face could be used as a meme.
me taking my 3d printer from 70mm/s to 90mm/s : this is a lot of spid and my printer its going to exploit This man go to 200mm/s to 800mm/s : yea its okey
Have a look at high speed CNC machining metal faster than what I am able to do with plastic . Those are scary! heavy heads, metal components.... failures can be pretty bad with those. The worse that happened to me was a broken heatbreak when my super volcano crashed the bed after a bad move.
@@MirageC Hehe yea this. When an axis take a nap in those machines due to a failure you can have metal-on-metal explosions from the extreme heat induced in the fraction of a second making a bit explode to pieces (or next to that launching the part being machined with scary speeds.)
I think the future of 3D printing if it is to go in the direction of quick prints (which honestly it probably won't, it'll go the convenience/quality direction most likely) it won't be with these filament type printers. It's like how hard drives became bottlenecked by the physical magnetic heads: slinging so much mass that it shakes the table is going to cause so much noise and take so much power. That said, this is absolutely awesome
Agreed! Light/magnetism/laser/electricity are probably what will form the 3d printer technology of tomorrow. That being said.... I need motors that will support 1000mm/s ! hahah
This is how fast 3D printers should run at. I know as technology advances this will become the norm and thank you for pushing the limits of what can be done. If 3D printing is to progress this sort of thing needs to happen. Be nice to see a benchy being made to see what the quality is like at the end of a high speed run.
Exactly my goal! Unlock the true potential of FDM printing. This technology can do a lot more. As for the benchy I have a go at it here: ruclips.net/video/ZZDuX6hcd28/видео.html But working on getting this number MUCH lower with the help of new firmware... Klipper :)
@@l3d-3dmaker58 How? With the same printer, I couldn't get more than 160mm/s, even when travelling. Any faster and all it will do is skip steps. At 120, the Y axis rattles to an extreme degree. I don't ever print faster than 65, or 40 for small models.
Hi Steven, All frame extrusions are 3030 T-Slots profiles. The X support member and Bed frame are 2020 profiles. You can see the required length according to your desired print area here: miragec79.github.io/HevORT/framecalculator.html
Amazing! I've not tried to do the math, but I wonder where the CPU performance of the controller being able to compute the kinemetrics and driving the steppers becomes the bottleneck? And when you're out of CPU cycles, how is does that manifest? A printer like that needs yellow and black hazard tape around out; you could put an eye out with that thing! :-)
I did not do the math until I red your comment ;) but here it is : - Duet 2 Wifi has a total step capacity of 300kHz - GT2 20 teeth pulley have a diameter of 12mm = 37.7mm circ - CoreXY staight moves are obtained using the two motors. 800mm/s is the the hypothenuse of the displacement. Each motor is supplying 565.7 mm/s. - I am using 400 steps/rev motors = - The motors are configured at 16x microstepping. (each physical step is divided in 16) Number of motor revs per sec = 565.7 mm/s / 37.7mm / rev = 15 revs / sec Number of steps per second = 15 revs/s * 400steps/rev * 16 = 96,034 steps/s for one motor. Total Steps for XY = 192,068 steps /sec The Z axis will be negligeable since it will be moving 0.1 mm each layer. Extruder = 16mm/s is required to achieve 0.1mm layer at 800mm/s on a 0.4mm nozzle. Using the E3D hemera @ 409 steps/mm this gives us 6544 steps/sec. Not a big deal. Total: 198,612 signals(steps) per second on a capacity of 300,000. I still have room :)
@@MirageC Damn that's epic! Like someone said, it looks like a timelapse, but in real time. Guess next step towards full madness is going with some NEMA 23s or bigger, on some TMC5160 to crank the amps all the way, and a super volcano at least.
@@MirageC Whoa looked there, damn that's insane (in a good way)! Definitely looking forward for what will come out of that. Would love to try something similar but I bet my whole budget for my machine is about the price of one of your servos so guess I'll have to wait till I'm well employed and have quite some spare income
I used to print at 0.15 mm layer height but after working out the numbers for my z layer height for my 4 start threads, it made sense to print at multiples of .04 mm, so .12,.16,.20 layer heights because the steppers don't have to maintain an in between step and there is no noise or z motor heat.
I think a larger printhead is needed with a longer heater barrel to melt more plastic. Reducing layer height just to go faster is useless because it will not produce any more output. Another option is a different nozzle design that can "brush" or pipe fully liquid material onto the layers and turn up the heat.
What step rates are you reaching up to? I am maxed out on my duet wifi 2 at around 110KHz around 200mm/s Do you have ideas to increase cpu load so to print faster?
I did not do anything specific. One thing that was limiting the max speed was low E accel and Jerk when using Pressure Advance. I now use M201 E3000 and M566 E3000.
How are you getting the 3 point probing speed between probing points to be so fast? Seems like no matter what I change I cannot get it to go very fast. All of my other movements are really fast at 25000mm/min but my 3 point is slow.
Look into you M558 command in your config.g file from your Duet. the T value is the travel between probing point. Oh! And I stronly advise that you add "B1" to that command line. This momentarily cuts the heating power while probing. Its been proven to make probing more accurate. Here is my M558 line as a reference: M558 P9 H15 F800 T30000 B1 ref: duet3d.dozuki.com/Wiki/Gcode#Section_M558_in_RepRapFirmware_2_x_and_earlier
If you're developing your own printers, I'd like to see the possibility with changing out the belts with ball screws and changing from steppers with small steps to the kinds of brushless motors seen as part of hobby RC aircraft. it can reduce ringing a good bit as the elasticity of the belts is the major cause of ringing, while also keeping step ratios similar to what they were on the belts, because geometry
What I have found to work best on glass surface is the printing material itself. I splash some acetone on the glass (that I cleaned with acetone) and rub a failed part on it. The acetone melts the material, forms a film of the printing material itself as acetone evaporates away. Quick to apply, disconnects clean once the glass cools, leaves glass like first layer surface.
I am truly considering building a HevORT for production use. I got two concerns tho, 1. What brand linear rails did you use? Genuine Hiwin ones or clones? 2. The Software, I've got a lot of experience with marlin based boards but repwrapfirmware is new to me. I really like the premise of the Duet boards but how hard is it to set everything up software wise? The bed leveling for example, is that hard to setup? Your design looks great man, it has everything I am looking for in a 3D Printer.
Hi Lenn, The rails on this printer are Genuine Hiwin rails. The software side of things is alot more simple with the Duet Serie Board than the marlin based boards. Some templates are being put together by the community. So it is expected to see an automated HevORT firware configurator in the near future.
yo tengo la prusa i3.....y para la adhesion,,,,,puse un cristal de 1.5mm y le aplico laca de pintura de los chinos,,,,eso me garantiza una adhesion espectacular,,,trabajado pla ,cama fria boquilla 0.2, velocidad 100,,buen video.
Really impressive! Did you try to print the same model but rotated at 45 degrees? Interesting to know how speeds and accelerations would scale. Will it be simply ~2 times slower?
You are right, steppers had a hard time keeping up with diagonals with those accels/jerks and speeds. I am now using Teknic ClearPath servo motors. Direction does not matter anymore :)
Hey, what is your machine's Z Height? AND If you don't mind, I'm asking as I need to swap a large format machine to 1204 (12mm) Ball screws X 800mm long each - with C7 Grade that's available at reasonable pricing, is there a good source for ballscrews you have used that don't have wobble along the length of them / cause issues mounting down the top bearing from being bent. Do you have issues like this? What Length are these? And do you know the link to the supplier? (It's hit or miss with these so if yours work I'd really hope I could get from the same source to not worry!) Thaks! Awesome Machine!
Hi. Is there a reason you chose the Herema Hotend over the Copperhead from slice engineering. Would not the extra heat capacity in the hotend assist in this record attempt?
Yes there is a reason. I did not know the copperhead or slice engineering until now. In this video I am using a copper Volcano hot end block with a tungsten carbide 0.4mm nozzle and a 65w heater cartridge. The Hemera is an extruder. The only gain I would get from more heat power is thicker layers. The xy motors are now the weak links for speed.
Quite infirmative. Now 3 years later I just bought a fruit of this speed race development a fast cheap printer Sovol SV07 with Clipper for only 220 $. Not core XY but hey its still fast and impresive for this price compared to suddenly obsolete popular Enders. 😉🤓👍
How high did you set the bed temperature on the last print? I am an absolute beginner and have massive issues getting perfect bed adhesion printing ABS on my cheaper Sunlu S8 with a glass bed. Tried glue, ramped the bed up to max heat (which sadly is 80 degrees only) and I‘m not printing fast, 50mm/s. Issue is thst I‘m printing a big Arduino enclosure (200x120x60mm) and that warping is pretty heavy.
Abs is tricky to get to stick on glass. The common procedure is to use an "ABS slurry". You take a piece of filament and mix it with acetone. This will create a mixture that you then apply to your bed. I personally switched all glue sticks a recipe to Magigoo. This stuff is amazing!
MirageC Thank you so much for the fast reply! I heard about this acetone trick and will give it a try next time I am printing with ABS, until then I will finish my prints in PLA which returns surprisingly great results on the Sunlu S8. I will let you know whether it worked out!
Strangely, PETG is great for speed printing, but has issues with stop and go (retraction and small moves). But on continuous moves and vase mode, I can go faster than with PLA.
Ici à Bruxelles, on apprécie beaucoup vos vidéos mais pas seulement vos vidéos et votre projet d'imprimante 3D haut de gamme, nous apprécions aussi vos trams (il n'y a pas de hasard, both are real bombers)!
nice man! if you're gonna go at those speeds, you'll need something like a Nova or a Supervolcano lol, but it's super nice to hear people starting to push the envelope! I absolutely love printing fast and think everyone should! it's certainly possible to make a commercial machine print 200mm/s out of the box
How does the fact that he prints with 657 mm/s and 8k acceleration with no problem makes you say that he needs a nova or supervolcano hotend? After all at that speeds, it's not about how fast you can melt plastic but how fast you can cool it down so it adheres and doesn't produce spaghetti.
Super Volcano will give me margin to increase layer height. It is sitting next to me... waiting eagerly for me to finish the design of its carriage. :)
@@MirageC I think the next issue will be the long heavy hotend flying around, because it is fastened at only one end with a weak heatbreak. Probably you will need some diagoal braces at the bottom in at least three directions. Made out of thin steel or stainless steel wires, which conducts heat badly and not heavy.
@@gabiold I am developing a solution for that matter. A piece of PCB phenolic with heat dissipation copper tracks that will link the bottom of the supper volcano back to the carriage :)
If this is done for fun, that's fine and that's fun. Other than that, during normal production it is useless to print at that speed as there will be quality/precision issues. And the moment there will be a small infill to print the vibrations will be so high that the printer will not stand it. With a properly modified delta it is possible to print a cylinder in vase mode at 1.000... and then? The point is that FDM printers are not made for speed. FDM 3D printers for fast production are the wrong tool. Like using a home 2D printer to print a banner. It would be better to invest time to find the best setup for the maximum quality/precision rather than pure speed.
Having tinkered and tuned my delta printer, I was proud of having it print very good quality items at 80 mm/s print speed with 1400 mm/s acceleration and 7 mm/s jerk. Then I keep coming back to this video and I'm thinking "hmmm I should be able to go faster, even with my inferior rigidity and slower board and crappy hotend". And now I'm running at 130 mm/s print speed with 2500 mm/s acceleration and 12 mm/s jerk. Print quality is still very acceptable and it'll save me tons of time when prototyping. And then I come back here again and I'm like, "hmmmm I should be able to go faster"... Well, today I'll try to go between 150-200 mm/s print speed with acceleration at 5000-7000 mm/s and jerk at 20-25 mm/s. We'll see when my printer will shit itself to pieces... THANKS for this video - it's very inspiring - and a coreXY is definitely on my "stuff to build" list...
What touchscreen are you using and what interface is that? It looks awesome. And also, I really like that satisfying sound the printer makes at the start of the print
I am using a rooted Amazon Fire 8 tablet to access the Duet internal Web Control. This control board has its own web server and you can control it from any device on your home network. Similar to Octopi or Klipper.
Pretty damn cool. Liked and subbed. Can I make a suggestion about improving your videos? Maybe try a script or bullet points. It gets a bit painful listening to anyone hunt for words, and some notes might help with that. I will definitely be keeping an eye on your channel to see what you do in the future!
It's good advice but entirely impractical for videos involving real-time commentary rather than post-facto. Beyond that, while word-hunting can get annoying, and sometimes very quickly for different people or in different contexts, use of a script can and will often make a production feel disingenuous and discomfiting to both producer and audience alike in a way that word-hunting rarely manages to approach. This becomes a significantly more likely outcome when the producer in question has a standing practice of real-time commentary, as here.
![COLOMBIA vs. CHILE [4-0] | RESUMEN | ELIMINATORIAS SUDAMERICANAS | FECHA 10](http://i.ytimg.com/vi/TDjN4W3yy1k/mqdefault.jpg)
"It is now printing at 200mm/s because it is the first layer, it will speed up..."
Let that sink in. I print my first layers at 20mm/s.
exactly my thoughts... now I am curious and want to see how fast I can print my first layer at ... maybe I can push my luck and do 30 mm/s? lol
My printer starts at 40mm/s but it’s a delta do it can do 100-200mm/s on regular layers
But this is incredible
If you had used a smaller nozzle could you go faster still
@@kendell8046 Wouldn't that actually result in a slower print speed? I don't know if you would be able to eject enough material through a small nozzle fast enough to get up to these speeds. Also, I think you would need a bit of a higher temperature to get the material to flow well enough through a smaller opening and not clog. But I could be wrong here, just a guess.
I think it would be fun to see a benchy or some other model speed run and see how fast you can get a successful print.
That's such a great idea, now i'm sad that this video didnt do that :(
@@Drakoman07 Because it would have came out like a turd.
Due to the short line segments in such a detailed model as benchy, I don't think it will accelerate all the way to the topspeed, so it wouldn't be a comparible result
There's not much difference in a model like a benchy, as the printer won't reach max speed on any segment. That's why the model he's used is a bunch of long straight lines, to give the printer a chance to accelerate to the target speed. It might be interesting to see a giant benchy though, scaled to the biggest side that would fit on the bed. I don't think it would be able to reach max speed but it should reach higher speeds than a 1:1 benchy.
@@suharsh96 Well then, if we're not going for quality here I'm sure I could crank out a benchy at 400 mm/s on my printer. Hell, i'll just tell it to shart out a bunch of filament in a pile at 800 mm/s and call it a benchy speedrun.
This is the 3D printing version of liquid nitrogen CPU overclocking, good stuff.
The thing is that these speeds are not accurate in the same way cpu clock speeds are read. There is no way to measure the speeds, sure anyone can order their printer to print at 1000 or 10,000 mm/s but the acceleration is still too low to achieve it. It certainly is moving very fast, just not 700mm/s fast.
In no way can you compare this to liquid nitrogen cooling! Because speeding ya printer up is actually a good idea
To be clear, this comment is a bit of a joke.
@@rikdenbreejen5230 XD
@@rikdenbreejen5230 It's just a playful analogy.
T Turner, my comment was a bit of a joke.
I'd love to see this beast printing a simple cylinder in vase mode. Should be able to reach even higher speeds without the cornering issue.
Great work.
Yeah I was thinking the same! Would be really cool.
But I guess the lesson for this video is that you need to work on the melting in the extruder. It would make more sense to print at 400m/s with 0.2 layer height than 800m/s at 0.1.
@@dejayrezme8617 unless you're a sucker for outside wall quality
Print my head!
the opposite is true.
that circluar vase has a few hundred thousand more corners then the print from this video
@sourand jaded
and it's not like the control software is going to give a damn about the angle of the bend.
infact it doesn't even know.
make a circle.
slice it.
and you get several thousand lines of code for tiny individual lines. (PER LAYER)
without ANY info on the angle of the next line.
and it's just coördinates.
G1 X0 Y0
G1 X0.01 Y0.0
G1 X0.02 Y0.01
G1 X0.03 Y0.01
G1 X0.04 Y0.02
excetra excetra excetra
and it is going to treat them as such.
you can tell it to go 800m/s all you want. it won't go faster then 50m/s (depending on the resolution of the circle)
you can then change play with the speed setting as much as you want. it's going to print it at the same speed each and every time.
however, if you change the jerk and acceleration, THEN suddenly it starts to go faster (or slower if you lower them)
if you have a 32bit motherboard and some special code in the firmware. it could look ahead in the g-code
and begin doing some math. (basicly turning all those G1 straight lines, into a G3 or G4 circle, then it suddenly does have the info of the angle on the corners, and it can decide to go faster because it doesn't need to slow down as much)
you can also install a plugin into the slicer. and as long as your firmware on the printer knows what G3 and G4 is. the slicer can just write (G3 with the coördinates of the centre point of the circle. the coördinates of the end point of the circle. and perhaps a radius) and it would be faster aswel.
and then you can print faster circles on an 8bit motherboard aswel.
next time you don't know jack shit about a subject. try to just ask a question. rather then pretend you know something and make a fool of yourself. or just shut up. either or.
learn how to program a lathe or mill with fanuc iso control before you come back.
(not heidenhein. that's the easy mode)
marlin and most 3d printer firmware is a childs toy compared to real cnc. and 3dprinting shortcommings show this well.
This just shows that 3d printing can progress to a speed that would print in acceptable times as the technology improves.
We are a long way off high speed prints for a while though - 3d printing is still pretty much in its infancy at this point..
I can only imagine the stress that all the motors are going through, this has to impact on their working life.. and the noise is something else entirely!!
Who knows how far we can go in the future? Who would have guessed 50 years ago that we could have computers the size of a matchbox and 3d printers making all kinds of objects, some even for medical applications?
Yeah, stuff of science fiction, but we will catch up..
Good video, well done..
Imagine in like 50-100 years when we're all printing live tissue with this sort of speed. Want a cat? 5 minutes, tops. Accidentally lobbed off a few fingers or your arm while working on the ol' hover-mobile? Go to Fiverr and have someone make you a new one.
Nikolai, you may just, but it's all entirely possible in the future.. if you watch today's sci-fi, it's a fair indication of our future..
Anyone remember the first 2d printers? These things were slow as heck, jammed all the time and required constant supervision - sounds familiar?
Probably we'll never achieve StarTrek replicator speed of 3d printing (because of that pesky thing called basic physics) but I am sure that we'll eventually achieve 3d prints that are done in minutes instead of hours and have acceptable quality. Maybe even full color 3d prints where printer would just melt together CMYKW (W=white since it cannot rely on the medium such as paper to supply that color) filaments in right proportions to get proper color.
@@UltimatePerfection in the movie "face off" (John Travolta/Nicolas Cage), there was an ear being 3d printed in a fashion very much like resin printers.
That film was 20years ago - who would have imagined that 3d ears and hearts are being made in today's medical industry?
You are right about 2d printers and the resemblance..
Who knows where well be in future years..
Just get servo motors, and then you Can optimize them to a ridiculous level
If you print a pentagram with 666m/s a portal to another world will open.
Take one print with infill man
doom guy: ah shit here we go again
666 meters/s seem kinda fast to me
I'll be impressed if you print anything at almost twice the speed of sound
u should use pentagram infill pattern!
Man,bless you for sharing your experience with the rest of the mortals. You should be the President of the 3D printing industry for pushing the limits beyond limits. Thank you and salutations from France.
yeah, we need more people like this! what are we, like 5 or 6 people? LOL this needs to take off, I got tired of waiting for prints, and I'm printing at 150 🤣
This is impressive. I’d love to see how it does laying down a complete bottom layer and maybe a more complicated part with some retraction. Glad you were recommended to me today. New sub.
it wouldn't it would fail 100% of the time
the industry is starting to reach the point where they need to start creating filament specifically for printers that run over 200mm/s
I wonder what the theoretical limit on speed is. At a certain point the filament would have to be near-liquid just to come out of the nozzle fast enough, and eventually it would just be a jet of molten filament coming out like a water jet knocking over anything already printed. At that point you might as well just stick a mold under it and call it an injection molder.
You'd need to focus on a longer heating chamber, so it has time to conduct heat to the centre of the filament before it reaches the tip.
@@DUIofPhysics i could see induction heating with a fluxed core like welding rod in the future as feed stock
@@nikolaivillitz6026 If we talking this to the extreme, we are talking about spray painting molten plastic!!
@@DUIofPhysics I'm melting more than 200g/h with my Nova hotend (47mm³/s or so) and I can tell you, neither specific filament or a ridículous long heatzone are the way to go, I can push that stupid amount of (relatively good) regular plástic (I think it's the same ingeo resin as prusament) and the Nova has a heatzone of about 18mm, sooo
It's just straight scary watching it move that fast.
you get used to it... after a couple dozens of prints🤣
ruclips.net/video/wkHjLJNt4oQ/видео.html
Dude, my hats off to you. I've got almost the same hardware as you but haven't added the volcano yet. I was stymied at 100 mmps by the melt rate. If that is what I can milk outta a stock volcano, then I can't wait to hotrod it. 🤯
volcano or Nova, depending on your budget, easy 300mm/s
This is completely insane. I am glad that you are one of the only persons that actually try and show what you can and what you can't do. This one of the most annoying parts of many DIY machine communities to me: Nobody wants to test or talk or show about the qualities of their machines. People literally say on their communitie's subreddit about their own machines "it's a 1000+ dollar machine, of course it prints very good". It is really, really frustrating.
Hope it is a $1000+ machine since I would think this is the cost in the end to build it.
Make it 2 of those 1000$ if you want to go with the ZR installation ;) studio.ruclips.net/user/videoaL7pEEHTTe4/edit/basic
@@MirageC :D (Just to clarify: I meant that people are telling that in their subreddits about their own machines - I edited my post to clarify that ^^. That's why I think it's great that you are testing and showing the limits and results from your machine here. I remember asking you in a previous video about that. You said you are planning on doing so. And you did it. And that's awesome.) - Your link for us viewers: ruclips.net/video/aL7pEEHTTe4/видео.html
no-trick-pony_lockpicking thank you for the honest word. That’s how I like to think. Nothing to hide here ;) just trying to move the 3D printing world one notch forward ( or confirm dead end! Hahah)
Incredible mate 😍😍😍 HevOrt is one of my favorite projects to follow and you never ceases to amaze with design, iterations, changes, and test videos that are absolutely outstanding !!! Incredible work 💪💪💪 !!
First time seeing this project. As an engineer I was already thinking from the title "how the heck could the tension belts and motors handle that speed". Your design solves that problem.
I'd imagine you'd run into severe stringing issues if you were printing something that required travel. Following👍
That is a very good test idea. Stringing at that speed. Never had an issue with stringing though. Pressure advance does most othe work. A slower retraction also help in order to avoid the fusion bath to break and let some material fall loose from the hot end.
Ha ha, was like watching a time lapse in real-time. 🤣
ruclips.net/video/wkHjLJNt4oQ/видео.html
this can print better and 10x faster than anything i could dream of, holy hell
WOW is that fast !!! 800 !!! Even 400 would be more than enough for me :-)))
yeah 400 is grerat and doesn't sound too abnormal xD
@@hyperhektor7733 yes think so, too. Have to try that with my printer 😂😂🔥
@DocH same thing that happens with computers: “this much RAM is so much, I’ll never use that much! This much hard drive, more than I’ll ever need! This fast of storage, faster than I’ll ever need (now moved to SSD)! This fast of CPU, faster than I’ll know how to use! This many cores, I won’t have a way to use all that!”
Performance inflation, same thing as lifestyle inflation: too much is never enough!
I love these Pushing The Envelope videos and can't wait for the technology to produce a $999 printer that can run 24/7 production at 400 mm/sec.
Cool project-I love how you built your printer, very close to how I want to eventually build my own at some point. The speeds you reach are incredible! You got a new subscriber. ^^
Great video. I like the way you new our to correct the problem each time. I have printers but not the high end like you got but I have learned something from it. Thanks and great technicians work.
A hell of a machine :D my frame is done today, I am also in your project group - awesome !
Thank you very much for bringing this all open source to the public. This machine in on my list!
I feel nervous when my printer is printing at 120 mm/s and your printer starts printing at 200mm/s...
whattt??
200 mm/s is the first layer after that it's up to 600-800 mm/s
Why?
I feel nervous printing at 40mm/s, and 65 is my never exceed speed.
@@user2C47 i print at 100mm/s.. and it prints great..
@@technicalbreakdown1484 My printer has a plastic frame and round idlers. It will rattle itself to (actual) death at 70mm/s.
We need more people working on this along with Top Down Resin printers... Right now my DLP printer is going through testing, at 10 cm per hour on the Z height with a build plate of 384 x 216 mm. Hoping to replace the LED in the projector to cut the time down to 35-45 minutes, the only problem then will be resin viscosity and wiper splashes.
Started watch at 2x speed: "Wow! Thats pretty fast!"
wow. incredible. you're even lugging around that heavy hemera and x-axis rail. I feel uncomfortable when my super-light bowden delta goes above 250. new sub.
As somebody who has to wait 47 hours for some prints, this both intrigues and sexually arouses me.
you would probbably shave about 1 hour off of that time with speeds like this.
you can tell your printer to print at the speed of sound. but if it doesn't have enoegh lenght in a straight line to ramp up to said speed. it will never get there.
i can tell my printer to print at 800mm/s but then print a cilinder. and it will only print that thing at about 100 maybe 150mm/s
why? because jerk and acceleration are limiting it's speed. because that circle is hundreds and hundreds of tiny straight lines that it keeps needing to slow down and accelerate for.
@@darkracer1252 say hellooooo to arc welder
Thanks, now I don't feel alone... Either a Hevort or a Voron build is very near in my future. I'm instantly obsessed.
@@darkracer1252 I think that's a great point. I have to ask, as I'm somewhat noobish, wouldn't the absolute maintainable speed (I want good print quality too) be directly dependent on the geometry of the part you are printing?
@@TheJacklwilliams yes it is. and as cris mawson says. arc welder helps (it's a plugin on cura that turns circles in circle commands rather then straight lines)
so the thousand or so lines of code for the short lines that make up a cirlce will be turned into a single code.
but what cris didn't understand is that that doesn't change what i said.
jerk and acceleration has nothing to do with the amount of code. arc welder only helps if you have an 8bit controller.
it's STILL thousands of tiny straight lines that all require it to take jerk and acceleration into account.
if you have "junction deviation" set up though it will slow down minimal because it takes the angle of the corner into the equasion.
so that's a jab at you cris mawson.
arc welder is solution to a slow cpu. not to slow jerk and acceleration.
I've never saw that amazing speed! It's something like: from dream to real in a eyes blink! :) My compliments.
Me watching this video where a printer prints with 800mm/s. My printer behind me with 50mm/s: Don't be worry I will finish in less than 9 hours.
3 days*
Very impressive for rapid prototyping good that you find the limits
I am doing 40-50 mm/s on my first layer an a Geeetech A10. 100 for infill, 80 for inner walls and I am probably killing some parts early with that...
Hey, what kind of software you use on tablet ? Iim using octoprint, but yours have more advanced settings.
The tablet is displaying the native Web Server from the DuetWifi board I am using. This control board will connect to your home network. You can then access it using your smart TV if you want :)
Can't wait to finish mine. Yours always amazes me and makes me hope that mine will be half as good as yours lol
Has it worked out yet?
@@joshanderson1019 Yes, just ghost printing tho... Still waiting on a hemera LOL.
@Thu Nell Ⓥ ^
Do post a video once you are done with it..
Wow, what a speed
Great update
Thanks for sharing 👍😀
I'd love to see a round vase being printed at that speed :D
I love seeing the zirc on that linear nut. Impressive machine. Congratulations.
Nice! At these speeds, you should really try an ODrive with BLDC motors. I am curious how fast speed could you achive with those.
Looking forward to the finished guide. I definitely want my own HevORT in my home!
If I blink the print will already be done
I can't figure out how your printer is so quiet at those speeds. Is tmc stealthchop working for you at these speeds? I (and many others) am battling with stealthchop being very noisy, to the point that spreadcycle mode is more quiet at >100mm/s. Which motors and pulleys are you using?
I imagine this in a sci-fi show where AI needs a spare part as fast as possible, and hacks the printer to the best of it's ability that AI determines.
Great video! You are pushing the limits of the industry!
My 3D printer:
*Spaghetti time*
My 3d printer:
*Skip steps without moving time*
*Shake itself to bits time*
@@user2C47 hahahaha
Going to look more into your printer if I decide to build it in the next couple of weeks definitely donating!
I want to see how extruder motor is spinning at those speeds 🤯.
I would be perfectly happy if I could reach 150, at 400 I was blown at 600 my face could be used as a meme.
OMG that is insane! All of it! Holy crap! My Tevo Tarantula would burst into flames and have a PLA hemorrhage just thinking about trying this!
me taking my 3d printer from 70mm/s to 90mm/s : this is a lot of spid and my printer its going to exploit
This man go to 200mm/s to 800mm/s : yea its okey
Nice printer. I love doing stuff for stuff's sake. Thanks for the video.
This gives me anxiety ! ! ! Imagine what a crash would do at those speed
Have a look at high speed CNC machining metal faster than what I am able to do with plastic . Those are scary! heavy heads, metal components.... failures can be pretty bad with those. The worse that happened to me was a broken heatbreak when my super volcano crashed the bed after a bad move.
@@MirageC Hehe yea this. When an axis take a nap in those machines due to a failure you can have metal-on-metal explosions from the extreme heat induced in the fraction of a second making a bit explode to pieces (or next to that launching the part being machined with scary speeds.)
I think the future of 3D printing if it is to go in the direction of quick prints (which honestly it probably won't, it'll go the convenience/quality direction most likely) it won't be with these filament type printers. It's like how hard drives became bottlenecked by the physical magnetic heads: slinging so much mass that it shakes the table is going to cause so much noise and take so much power. That said, this is absolutely awesome
Agreed! Light/magnetism/laser/electricity are probably what will form the 3d printer technology of tomorrow. That being said.... I need motors that will support 1000mm/s ! hahah
If you need to go faster without skipping steps, try what is used in slot machines with large reels.
what is used?
This is how fast 3D printers should run at. I know as technology advances this will become the norm and thank you for pushing the limits of what can be done.
If 3D printing is to progress this sort of thing needs to happen.
Be nice to see a benchy being made to see what the quality is like at the end of a high speed run.
Exactly my goal! Unlock the true potential of FDM printing. This technology can do a lot more.
As for the benchy I have a go at it here: ruclips.net/video/ZZDuX6hcd28/видео.html
But working on getting this number MUCH lower with the help of new firmware... Klipper :)
I can do the same(fail) at 60 mm/s
hahahhahhahha bro
Hahaha I can made it at 20mm/s, haha it's so sad
lol I'm doing 300 on an anet a8 over here🤣 that's what I call sketchy
@@l3d-3dmaker58 did you replace the frame with a metal one?
@@l3d-3dmaker58 How? With the same printer, I couldn't get more than 160mm/s, even when travelling. Any faster and all it will do is skip steps. At 120, the Y axis rattles to an extreme degree. I don't ever print faster than 65, or 40 for small models.
Cool. A speed test is an extruder test though. Acceleration is a printer test.
What if people had competitions to see how fast they could make their 3D printers print things
YEEEEESSSSSS
@@Silverdev2482 Benchy world records are a thing now!
@@Omlet221 YAAAAAAAAAAAAAAAAAAAA
Really puts the Rapid in Rapid prototyping.
Simple solution is to install a “flux capacitor”
then you finish the print before even starting?
"I gonna start the print now, but it was finished 30 years ago"
That's awesome! One step closer to replicators. Keep up the awesome work mate. 👍
Now print an actual complex and I would be impressed.
Yes boss! ;)
That was his main goal, impressing you.
adisharr sure sure
I'm impressed that this machine can handle continuously operating at speeds this fast at all while maintaining such precision. Duet > Marlin
I mean this is pretty impressive as is
What extrusion are you using for that? It almost looks like 8020 15 series and I have a bunch of spare aluminum in that size....
Hi Steven, All frame extrusions are 3030 T-Slots profiles. The X support member and Bed frame are 2020 profiles. You can see the required length according to your desired print area here: miragec79.github.io/HevORT/framecalculator.html
@@MirageC darn, was really hoping you used the 15 series lol
Amazing! I've not tried to do the math, but I wonder where the CPU performance of the controller being able to compute the kinemetrics and driving the steppers becomes the bottleneck? And when you're out of CPU cycles, how is does that manifest?
A printer like that needs yellow and black hazard tape around out; you could put an eye out with that thing! :-)
I did not do the math until I red your comment ;) but here it is :
- Duet 2 Wifi has a total step capacity of 300kHz
- GT2 20 teeth pulley have a diameter of 12mm = 37.7mm circ
- CoreXY staight moves are obtained using the two motors. 800mm/s is the the hypothenuse of the displacement. Each motor is supplying 565.7 mm/s.
- I am using 400 steps/rev motors =
- The motors are configured at 16x microstepping. (each physical step is divided in 16)
Number of motor revs per sec = 565.7 mm/s / 37.7mm / rev = 15 revs / sec
Number of steps per second = 15 revs/s * 400steps/rev * 16 = 96,034 steps/s for one motor.
Total Steps for XY = 192,068 steps /sec
The Z axis will be negligeable since it will be moving 0.1 mm each layer.
Extruder = 16mm/s is required to achieve 0.1mm layer at 800mm/s on a 0.4mm nozzle. Using the E3D hemera @ 409 steps/mm this gives us 6544 steps/sec. Not a big deal.
Total: 198,612 signals(steps) per second on a capacity of 300,000.
I still have room :)
@@MirageC Damn that's epic! Like someone said, it looks like a timelapse, but in real time.
Guess next step towards full madness is going with some NEMA 23s or bigger, on some TMC5160 to crank the amps all the way, and a super volcano at least.
@@Kalvinjj Thanks! Next step is already under development.. SERVOS! :)
ruclips.net/video/m6DoKoESPdg/видео.html
@@MirageC Whoa looked there, damn that's insane (in a good way)!
Definitely looking forward for what will come out of that.
Would love to try something similar but I bet my whole budget for my machine is about the price of one of your servos so guess I'll have to wait till I'm well employed and have quite some spare income
I used to print at 0.15 mm layer height but after working out the numbers for my z layer height for my 4 start threads, it made sense to print at multiples of .04 mm, so .12,.16,.20 layer heights because the steppers don't have to maintain an in between step and there is no noise or z motor heat.
Disclaimer: the speed video reproduction is 1x.. :V
ruclips.net/video/wkHjLJNt4oQ/видео.html
Amazing... Keep up the good work... Love to see polished results...
@MirageC What printer is this? Did you build it yourself or is it some commercially available, heavily modded printer?
This is my own design. You can find more at www.hevort.com :)
MirageC is there any way that we can get in touch? Mail... Messenger... Whatsapp?
I think a larger printhead is needed with a longer heater barrel to melt more plastic. Reducing layer height just to go faster is useless because it will not produce any more output. Another option is a different nozzle design that can "brush" or pipe fully liquid material onto the layers and turn up the heat.
What step rates are you reaching up to? I am maxed out on my duet wifi 2 at around 110KHz around 200mm/s
Do you have ideas to increase cpu load so to print faster?
I did not do anything specific. One thing that was limiting the max speed was low E accel and Jerk when using Pressure Advance. I now use M201 E3000 and M566 E3000.
Wow I've never see anything like this before fantastic. 👍
How are you getting the 3 point probing speed between probing points to be so fast? Seems like no matter what I change I cannot get it to go very fast. All of my other movements are really fast at 25000mm/min but my 3 point is slow.
Look into you M558 command in your config.g file from your Duet. the T value is the travel between probing point. Oh! And I stronly advise that you add "B1" to that command line. This momentarily cuts the heating power while probing. Its been proven to make probing more accurate. Here is my M558 line as a reference:
M558 P9 H15 F800 T30000 B1
ref: duet3d.dozuki.com/Wiki/Gcode#Section_M558_in_RepRapFirmware_2_x_and_earlier
Insane! I love to see how much you can push these things
If you're developing your own printers, I'd like to see the possibility with changing out the belts with ball screws and changing from steppers with small steps to the kinds of brushless motors seen as part of hobby RC aircraft. it can reduce ringing a good bit as the elasticity of the belts is the major cause of ringing, while also keeping step ratios similar to what they were on the belts, because geometry
Scottie turn up the inertia dampers, dam it Jim, its all she got. Cool best of luck to you, interesting you have the need for speed, love it. Cheers!
Nice build. Awesome speed.
Seeing this now lol pretty sure you were well over 1200mm/s recently. It's amazing to see the progress.
What I have found to work best on glass surface is the printing material itself. I splash some acetone on the glass (that I cleaned with acetone) and rub a failed part on it. The acetone melts the material, forms a film of the printing material itself as acetone evaporates away. Quick to apply, disconnects clean once the glass cools, leaves glass like first layer surface.
Yes, this works great with ABS, but PLA and PETG dont disolve well or at all with acetone.
I am truly considering building a HevORT for production use. I got two concerns tho,
1. What brand linear rails did you use? Genuine Hiwin ones or clones?
2. The Software, I've got a lot of experience with marlin based boards but repwrapfirmware is new to me. I really like the premise of the Duet boards but how hard is it to set everything up software wise?
The bed leveling for example, is that hard to setup?
Your design looks great man, it has everything I am looking for in a 3D Printer.
Hi Lenn,
The rails on this printer are Genuine Hiwin rails. The software side of things is alot more simple with the Duet Serie Board than the marlin based boards. Some templates are being put together by the community. So it is expected to see an automated HevORT firware configurator in the near future.
@@MirageC Thanks for your reply! Good to hear!
Sourcing the components right now. Sadly the rails become really expensive after ~30cm.
yo tengo la prusa i3.....y para la adhesion,,,,,puse un cristal de 1.5mm y le aplico laca de pintura de los chinos,,,,eso me garantiza una adhesion espectacular,,,trabajado pla ,cama fria boquilla 0.2, velocidad 100,,buen video.
Lol, Im gonna try these settings on my printer🤣
Congrats and this is amazing stuff!!
Would love to see a complete print of a pint sized vase at 800mm/s. This video is awesome. Truly in awe.
Thank you! Planning to do more testing with Vases and benchy soon. Some modification being done as we speak.
Really impressive! Did you try to print the same model but rotated at 45 degrees? Interesting to know how speeds and accelerations would scale. Will it be simply ~2 times slower?
You are right, steppers had a hard time keeping up with diagonals with those accels/jerks and speeds. I am now using Teknic ClearPath servo motors. Direction does not matter anymore :)
Hey, what is your machine's Z Height? AND If you don't mind, I'm asking as I need to swap a large format machine to 1204 (12mm) Ball screws X 800mm long each - with C7 Grade that's available at reasonable pricing, is there a good source for ballscrews you have used that don't have wobble along the length of them / cause issues mounting down the top bearing from being bent. Do you have issues like this? What Length are these? And do you know the link to the supplier? (It's hit or miss with these so if yours work I'd really hope I could get from the same source to not worry!) Thaks! Awesome Machine!
Sir this is very awesome print
Hi. Is there a reason you chose the Herema Hotend over the Copperhead from slice engineering. Would not the extra heat capacity in the hotend assist in this record attempt?
Yes there is a reason. I did not know the copperhead or slice engineering until now. In this video I am using a copper Volcano hot end block with a tungsten carbide 0.4mm nozzle and a 65w heater cartridge. The Hemera is an extruder.
The only gain I would get from more heat power is thicker layers. The xy motors are now the weak links for speed.
Quite infirmative. Now 3 years later I just bought a fruit of this speed race development a fast cheap printer Sovol SV07 with Clipper for only 220 $. Not core XY but hey its still fast and impresive for this price compared to suddenly obsolete popular Enders. 😉🤓👍
How high did you set the bed temperature on the last print? I am an absolute beginner and have massive issues getting perfect bed adhesion printing ABS on my cheaper Sunlu S8 with a glass bed. Tried glue, ramped the bed up to max heat (which sadly is 80 degrees only) and I‘m not printing fast, 50mm/s. Issue is thst I‘m printing a big Arduino enclosure (200x120x60mm) and that warping is pretty heavy.
Abs is tricky to get to stick on glass. The common procedure is to use an "ABS slurry". You take a piece of filament and mix it with acetone. This will create a mixture that you then apply to your bed. I personally switched all glue sticks a recipe to Magigoo. This stuff is amazing!
MirageC Thank you so much for the fast reply! I heard about this acetone trick and will give it a try next time I am printing with ABS, until then I will finish my prints in PLA which returns surprisingly great results on the Sunlu S8.
I will let you know whether it worked out!
For printing fast, are you able to print PETG at higher speeds, or are you limited to specific materials at such high speeds?
Strangely, PETG is great for speed printing, but has issues with stop and go (retraction and small moves). But on continuous moves and vase mode, I can go faster than with PLA.
watching this while my Ender 5 pro is printing at a snail's pace... The speed difference is amazing
Thank you! I like these videos much more than fancy videos "10 things how to print" with millions of views.
Congratulations, good job, I want to build one too. You gave me a good inspiration.
Ici à Bruxelles, on apprécie beaucoup vos vidéos mais pas seulement vos vidéos et votre projet d'imprimante 3D haut de gamme, nous apprécions aussi vos trams (il n'y a pas de hasard, both are real bombers)!
Phlupke et ici à Montréal on vous remercie beaucoup pour votre mot d’encouragement:)
How much infill for the 3d printed parts please? I looked at how to print your parts but wasn't sure. Thank you!
30% infill with 5 bottom and top layers. And 3-4 walls. That is what I use. ;)
nice man! if you're gonna go at those speeds, you'll need something like a Nova or a Supervolcano lol, but it's super nice to hear people starting to push the envelope! I absolutely love printing fast and think everyone should! it's certainly possible to make a commercial machine print 200mm/s out of the box
How does the fact that he prints with 657 mm/s and 8k acceleration with no problem makes you say that he needs a nova or supervolcano hotend? After all at that speeds, it's not about how fast you can melt plastic but how fast you can cool it down so it adheres and doesn't produce spaghetti.
Super Volcano will give me margin to increase layer height. It is sitting next to me... waiting eagerly for me to finish the design of its carriage. :)
@@MirageC I think the next issue will be the long heavy hotend flying around, because it is fastened at only one end with a weak heatbreak. Probably you will need some diagoal braces at the bottom in at least three directions. Made out of thin steel or stainless steel wires, which conducts heat badly and not heavy.
@@gabiold I am developing a solution for that matter. A piece of PCB phenolic with heat dissipation copper tracks that will link the bottom of the supper volcano back to the carriage :)
@@MirageC Oh, cool! We eagerly waiting to see it in action. 😉
This makes my delta look like a tall snail! Amazing stuff! Make a part next time please!
StreetArtistsOfTheWorld I will. :)
what delta you have? I've seen a Klssel linear go 1000mm/s
@@l3d-3dmaker58 I have a anycubic kossel linear plus, I upgraded to a e3d nozzle (heatsink, heater etc) and right now I run it at around 60-70mm/s
I have a Prusa i3, but still felt i learnt useful info on diagnosing print issues from your video. Amazing speeds as well!
If this is done for fun, that's fine and that's fun. Other than that, during normal production it is useless to print at that speed as there will be quality/precision issues. And the moment there will be a small infill to print the vibrations will be so high that the printer will not stand it. With a properly modified delta it is possible to print a cylinder in vase mode at 1.000... and then? The point is that FDM printers are not made for speed. FDM 3D printers for fast production are the wrong tool. Like using a home 2D printer to print a banner. It would be better to invest time to find the best setup for the maximum quality/precision rather than pure speed.
Having tinkered and tuned my delta printer, I was proud of having it print very good quality items at 80 mm/s print speed with 1400 mm/s acceleration and 7 mm/s jerk. Then I keep coming back to this video and I'm thinking "hmmm I should be able to go faster, even with my inferior rigidity and slower board and crappy hotend".
And now I'm running at 130 mm/s print speed with 2500 mm/s acceleration and 12 mm/s jerk. Print quality is still very acceptable and it'll save me tons of time when prototyping.
And then I come back here again and I'm like, "hmmmm I should be able to go faster"...
Well, today I'll try to go between 150-200 mm/s print speed with acceleration at 5000-7000 mm/s and jerk at 20-25 mm/s. We'll see when my printer will shit itself to pieces...
THANKS for this video - it's very inspiring - and a coreXY is definitely on my "stuff to build" list...
love seeing that printhead flying across the bed!
me trying to print at 100 mm/s on a tronxy XY-2 pro: my whole printer: fuck this shit I'm out
What touchscreen are you using and what interface is that? It looks awesome. And also, I really like that satisfying sound the printer makes at the start of the print
I am using a rooted Amazon Fire 8 tablet to access the Duet internal Web Control. This control board has its own web server and you can control it from any device on your home network. Similar to Octopi or Klipper.
Pretty damn cool. Liked and subbed.
Can I make a suggestion about improving your videos? Maybe try a script or bullet points. It gets a bit painful listening to anyone hunt for words, and some notes might help with that.
I will definitely be keeping an eye on your channel to see what you do in the future!
Thank you for the constructive feedback. :)
It's good advice but entirely impractical for videos involving real-time commentary rather than post-facto. Beyond that, while word-hunting can get annoying, and sometimes very quickly for different people or in different contexts, use of a script can and will often make a production feel disingenuous and discomfiting to both producer and audience alike in a way that word-hunting rarely manages to approach. This becomes a significantly more likely outcome when the producer in question has a standing practice of real-time commentary, as here.