Is a lowered belt printer BETTER?
Vložit
- čas přidán 8. 04. 2021
- Become smarter in 5 minutes by signing up for free today: cen.yt/mbthomassanladerer3 - Thanks to Morning Brew for sponsoring today’s video.
⬇️ 3D models, filaments, printer links⬇️
Models used
Armel the polar paladin www.prusaprinters.org/prints/...
3D Benchy www.prusaprinters.org/prints/...
CR-30 conversion kit www.prusaprinters.org/prints/...
Get the CR-30
Via the Kickstarter page go.toms3d.org/CR30
Directly from Creality go.toms3d.org/CR30CR
Filaments used
No-label turquoise PLA
Prusament Urban Grey PETG go.toms3d.org/Prusament
DAS FILAMENT Transluzent Neongrün PETG www.dasfilament.de/filament-s...
Product links are affiliate links - I may earn a commission on qualifying purchases (at no extra cost to you)
🎥 All my video gear toms3d.org/my-gear
I use Epidemic Sound, sign up for a 30-day free trial here share.epidemicsound.com/MadeWi...
🎧 Check out the Meltzone Podcast (with CNC Kitchen)! / @themeltzone
👐 Enjoying the videos? Support my work on Patreon! / toms3dp - Věda a technologie
"Two areas specifically that are causing problems; that is top surfaces and bottom surfaces, and the infill itself". Other than that, no problem :D
Good walls then?
That's exactly what my first thought was
@@colinyoung3685 shortly followed by this
would love to see like a traditional cr10 with the belt, so when the print is done it moves out of the way and next print starts, cut out the whole 45 print angle dilemmas
That’s not the reason for this belt printer they wont to b able to print for example a one meter sword with this machine.
If you want to print several things in a row this is not the way to go.
@@felixtrefzer3893 out of curiosity, what would be your ideal solution if you were looking to print a large number of individual parts with minimal downtime?
@@WeAreChecking for me something like Thomas voron it’s Great to do that. But it’s a littlebit over engendered.
For less down time the simplest way you can go is to push the prints off bed by the printer itself.
@@felixtrefzer3893 that's not necessarily true. There could be a sequence that would allow part to print, then move the belt, continue printing.
@@dominick253 it would still have to be sliced diagonally or use non-planar 3D printing. Extruders still have a width to them, so you can’t just do sections.
Flipping the CR-30 by 45° to get a "normal CoreXY" blew my mind
it's almost like it's half a cube.. 😋
"This was actually a lot easier than I though it would be" followed by hours of troubleshooting... I know this feeling too well.
CR-30: The LowRider Edition! Awesome name.
Cool one Tom! I love all that insight as well as the quick glimpses at the failed tests. Experimenting with changes and seeing how far they can be pushed is something I rarely see. Awesome!
I love your investigation and the thought/analysis that went into this project. Keep it your ideas flowing.
I just want that belt as a bed for my ender 5 and use it for reliable parts ejection.
Honestly, I don't need an infinite y axis. The 45° angle looks horrible too
Really good idea. 6cm isn't really all that low for a large amount of prints. It's a worthy trade off in return for the advantages you were seeking.
You monster! Just kidding😅 I love it- moar 3DPrintMill hacks!
You can feel the fury building up
Naomi be preparing all her cyborg ninja tech weapons & armour to pay Thomas a visit
Was low key hoping to find you in the comments! Also, it seems he has replied to 3D Printing Nerd's comment, but not yours...hmmm. maybe he is afraid XD. As an aside, I cant wait to get one of these belt printers in the future, I might be more inclined to mess around with cosplay :D.
Naomi ‘SexyCyborg’ Wu, do you think a PEI coated belt would help with his adhesion issues? He is using petg.
Is there some reason why it the belt can't be tilted the other way and the nozzle rotated 90 degrees to make a regular X-Y-Z machine with the belt being the Y? I would imagine that this was tried first, then the belt was turned 45 degrees for some practical reason. Would you mind sharing that reason and why having the nozzle at 90 degrees to the belt was a problem to solve?
This video is really cool! Great build. I love the musical choices.
Such a clever idea, and excellent execution! Can't wait to see "adjustable angle" belt printers become more common!
I was thinking the same thing as the video went on. Add 2 more stepper motors and a hinge and some shenanigans and you've got yourself a masterpiece.
@@Justkyrios Then we could also have a go at nonplanar printing. The only thing that remains is slicer support and proper bed materials for adhesion.
@@Tubaii25unfortunately all is missing is patents expiring...
Adjustable angle printers -> patented
i3 printers with conveyor belts for Y -> patented
Non-planar printing -> patented
I do understand why patents exist and their functions, but they are really holding back the full potential of 3D printing, imho.
Both Tom and Stefan have been going through this topic on their podcast, several times...
@@andy_liga Yeah.. forgot about the patents side of these things 😔
but if the community just started building things without selling them would a patent really matter? 🤔
Thanks for testing and sharing this
Hey Tom, It's a while since I've watched one of your videos. Your editing and general production values have become way better. Just thought I'd let you know - as a sort of virtual pat on the back.
I wonder if you used Kiri:Moto for the slicer would have made a difference?
I'll need to try Kiri:Moto with the CR-30 some day... Ideamaker, too!
@@MadeWithLayers Hmmm... What about a second extruder at the back--perpendicular to the belt--that lays down a thin, hot layer for the main extruder to lay the model on?
Once my CR-30 arrives, I'd really like to try Proper Printing's "Moving Portal Mod" to basically turn it into an ABP. 'Course, such a thing can't be commercialized til Makerbot's patent runs out in 2030, but a lights-out 3D printing work cell sure would be nice!
@@KeithOlson i was thinking the same thing
You should increase your top layer and bottom layer to 7 layers
Interesting concept.... what if the nozzle itself was ground at a 15 degree angle? You would have to compensate for increased line width due to the elliptical hole, but it would be a fun experiment.
Hmmm... I was about to say it would also be possible to design an angled nozzle, but then the problem would likely be getting it to properly tighten at *just* the right angle
@@jasonking7570 you might also run into flow issues with an angle that sharp, you'd probably have to angle the whole hotend, and I don't know if that infringes on the patent that these angled belt printers avoid or not.
@@Redeemed1337 It really doesn't matter if it infringed on the patent since that only applies to products you're selling. You're free to infringe on all the patents you want at home for your own use.
@@xKatjaxPurrsx In the US, you can indeed infringe on the patent by just making something for use at home. Obviously, this isn't enforceable, but the moment you put that information on the internet... Well. Have fun.
@@Redeemed1337 citation needed...
It was the mint blue filament! I couldn’t get that filament to not curl on my regular ender 3 pro! It was some sort of fila cube pla + 2nd gen or something that didnt require the be to be heated?
Thomas I just found this video so informative. I've got one of the 1st CR30s commercially released and found that I have had very low adhesion compared to the other CR30s videos I've watched. I have 3 other fdm printers and this adhesion issue made no sense what so ever until your video. Could you or Naomi Wu tell me a bit about the different belts as I believe that after all my sttuggles that I have an early low adhesion version. Thanks Dave.
I love your new battle bot build mate.
Would doing a mash up of the CR-30 and an Ender do any better?
The idea would be to have it print a standard model like the Ender normally would, but then have the belt rotate until the model falls off, then start another one, repeat...
Would be a lot trickier but also pretty neat: software adjustable angle so you can get the perfect first layer
Best parts:
(1) Build montage and music
(2) "Heated bed. Hot fire. Cooled bed. Snowflake."
(3) Tom's shop. Seriously, he could build just about anything there. I need to clean the garage and fix the door...
(4) Modding a printer without chopping the stock aluminum extrusions.
Could you make a 45 degree nozzle. I feel like that would fix everything
That probably won't work. If you angle the nozzle you'd get better bed adhesion but your layer adhesion will suffer the same amount as the bed adhesion improves.
This style of video is really cool
Could you tilt the model 15 degrees in a normal slicer? If you don't need support
Hey Thomas Sanladerer ! Do you think it will have an effect if one would grind the nozzle in the same angle like the printer?
keep the 45 degrees and just make a new mount for the hotend that makes the nozzle 0 degrees to the bed. I did this and it made it print the same as a regular printer but with infinite Z.
Exactly what I was going to suggest.
Tool-head clearance becomes an issue.
I wonder what kind of engineering it would take to design a new nozzle so that the entire machine is at it's original 45 degree configuration, however only the tip of the nozzle is angled 45 degrees down so it extrudes straight down on to the belt.
But then it would be angled wrong for when its printing the rest of the part.
What was the sound track during the build montage? Reminds me of a few 90's bands.
One question, what if you would use the belt as the Y-axis (the bed) and the use this to mass produce parts? (so instead of using this printer to produce meter long parts, just use it to produce repeated prints?
I've seen people use this printer to produce mass amounts of clips and things like that so its doable
Will non planer work to resolve the slicer angle issue ?
I've only used Ideamaker with my CR-30, but I've never had any adhesion issues at all. If anything, it's farther towards sticking too well. This is using PLA; I haven't tried anything with PETG yet. Only real issue I've had is the profile I started with prints infill almost like spider web. It's only barely capable of holding up the top surface and the models feel really weak. Changing patterns and density helped out with that to a large extent. I do still screw up about every other print as I test part orientation and then have to reposition on the printer to better align with the printing pattern. I assume that eventually I'll get the hang of it. I went through the same learning process when I got my first SLA printer, having to learn a new set of printing rules that were largely the opposite of the FDM rules I was familiar with.
Oh my god, i was jus trying to figure out how to calibrate e steps so i watched your video from 2014 and the difference between then and now looks like 2 different channels
So, I take it the reason that a delta isn't mounted onto one of these to print perpendicular on the heated side is some sort of patent, right?
What about Core-XYZ printer (Voron 2?) with belted bed? I think that might be a way to print reliably (bed would not move during printing) and continuously. And maybe with a bit of non planar printing you might even print those very large models.
This video kinda proves that Prusa was right and this famous print mill is just o piece of...not worthy of owning expensive piece...
Anyway thanks for your video Tom, are there gonna be more filament extruder videos?
What about making a mount for the hot end that holds it at a slight angle so the print head is at 90 deg to the bed
Then the head isn't perpendicular to the layer anymore.
You just convinced me I don't want a belt printer. The local micro center just got one. I am a unique proponent of printing at angles to strategically reduce bad overhangs. Eg a hollow cube ..
What if you had the infill as part of the model? Rather than going with auto-generated infill, have it pre-existing, so that when sliced, it slices it as if it were a wall, making them line up as naturally as the walls do.
why couldn't you have a regular xyz printer like an ender 3 and just have the belt move back and forth as if it were the bed and then at the end of the file is can move forward for the entire length of the bed to clear it and bring a fresh printing space?
This printer is so under engineered! They could have easily made a tilt feature on the hotend for the beginning lip of the print where it angles off after it gets started. I also feel like you should have made a variable height setup rather than cutting the extrusions like you did. You could have easily put hinges on the back side, and then some type of slider with linear rods on the front side to make it adjustable depending on the model.
*Awesome montage!*
Good vid, I’ve wanted to get a 3D printer for a while, but had no need till now. I do leather crafting and want to be able to make my own stamps/makers mark.
Is there a printer that you could recommend for a beginner? Would need to be able to print a material that can be whacked with a maul or weighed down on leather to stamp an impression.
i have actually tried using pla to do just this. a high infill print and pressure makes great stamps. i wouldn't wack them with a maul though
Another way to achieve a similar thing, would be to just tilt the hotend to be more vertical. If you then can also make the bed move forward and back during a layer, the effect would be the exact same.
That way you don't loose an z height.
If you just adjust Z to sin(offset angle)*y + normal z
That way, you don't have to change the printer as much, and it could even be adjustable.
I don't know though if the Z-bed bed mechanism is robust enough for this, as as its now a true 3D process not 2.5D
Are there any belt based printers that have no tilt? so they print normally but can just dump the part off and start again?
i don't think so, partly because of that makerbot/stratasys patent about it
One thing that needs to be done in the software to prevent those internal issues is to actually skew the model before slicing. The problem I had when designing the software was that wasn't a feature available to me back in the day. I wound up just rotating the model to the inverse belt angle... slicing it, then adding the Z shift to compensate for the offset. But if the model were already skewed to match the offset of the belt, then no post changes would be necessary and the internal "additional" layers would be in the correct positions.
Would it be possible to show how to do the skewing in CAD before slicing?
@@shamusobi2748 Yes... or just a pre-processor on the STL.
Isn't that exactly what today's belt slicers are already doing?
It's the order of precedence that is wrong... those features need to be identified before the skew... but generated after it... so they're generated at the proper offset (angle) from the layer above.
Cura wasn't available when I wrote the first post processors... but it should be easy to modify it to do it in the correct order. Post processing won't do it correctly.
Could you use the xy system from the cr30 on an ender 5 pro?
Maybe a stupid question, instead of changing the angle to 15 degrees this way why not keeping the gantry in same config and create a plate for the head that change the angle of the head vs the plate? This can bring the angle to more vertical without reducing the print size. Even better this cn be made potentially adjustable with a small stepper and gaining possibility to adjust it by the slicer depending on the best angle to print. If you add a limited possibility to move also the belt back and forth you can add additional “axes” capabilities. Or I’m missing something?
Needs an extra axis on the print head so it can rotate from parallel to the bed to 45 degrees. Not sure if its viable but it would solve some issues.
With the benchys losing adhesion at a consistent height, did you check to see if that is where they transition from the heated bed to the cold portion of the belt?
Good point... I'd think they should still be on the heated plate at that point, but I'm going to have to check
He did mention that as a general problem at the beginning of the video. But the bigger issue probably is the one he explained. At a higher bed angle you get a much higher number of overlapping extrusions that re-heat the previous layer and push it further into the belt surface. I would've expected serious warping/lifting issues on that last print if the the uneven bed temperatures were the main culprit.
This might make for an interesting production machine given the right print. It's basically a conveyer belt, you can go on the the next print as you dump the other one off into a bin/transport belt etc
And I actually felt my face change expressions when I heard that noise @ 7:27 lol That sounded horrific!
could one make a regular printer with just a belt instead of a bed, like for making lots of parts in a row on a ender 3 style printer, and it just movers the belt to a "new build area" between each print?
You can. But we don't need belt bed.
Why not mount the extruder on a 45° bracket, then it's at 90° to the bed/belt?
Use raft on the bed surface. It will Turn out really good. More filament but it works
I wonder if close to 100% (solid or thin objects) or 0% infill (hollow objects) works best with this configuration. Its kind of a waste if material but for some projects it might be best
Congrats on your channel for being so successful both on informing and entertaining us 3d printer lovers
Before you did all of that, did you at least try raising the front edge of the printer a couple / few inches just to see what would happen if you oriented it more towards the back?
You could have just designed a duel fan, duct and have each ducked cover each area, maybe one with higher air pressure to the extreme overhang on the far side of the nozzle, very minor firmware tweaks to PWM at best
I wonder if you would be able to design something to tilt the carriage that last 15 degrees so that it printed with the nozzle directly contacting the belt, or do the same thing in the original design. That way, bed adhesion wouldnt be an issue.
For the infill, wouldn't it work better, if the infill was sliced straight up as usual and only the shells being slanted?
what's the name of the song during the building phase????
@Thomas Sanladerer
have you tried with 30° ? i think 15° is to drastic change ...
wouldn't it make more sense to just angle the toolhead rather than change the tilt of the X/Z gantry?
edit: while I am thinking about it, it might actually make even more sense to rotate the gantry the other direction(closer to vertical) but I'd have to play with the idea in CAD for a while to conceptualize it further
How about instead of altering the whole frame to get from 45 to 15, you had only altered the print head mount to only rotate the print head by those 35 degrees?
That’s a great idea. I wonder if that’s do-able.
What was the song name during the build montage?
Hey Tom! Have you seen the crazy print quality coming off of the BLV MGNs? I'd be delighted if you covered those beasts!
Would be interesting how other angles would affect the print.
An angle sanded nozzle to be more paralell to the bed, would that do anything?
I'm not a lawyer but that might breach the patent, since it deposits at a 90 degree angle, even though the rest of the printhead would be at 45 degrees. Also, sanding a tube at 45 degrees leads to an oval cross section of the opening, which means that the "dots of plastic" would be wider in one direction than the other. That might also influence print quality.
cool idea
I think that you could sand nozzle extreme to get a parallel extrusion with bed. Even some company could develope a nozzle with this specific feature
You mean sanding the tip at a 45 degree angle? I think you could, but the nozzle opening would be oval, not circular. That might influence the quality of the print, since each "dot of plastic" deposited would be wider in one direction than the other.
Also, there is a patent covering 90 degree angle belt printers and I think depositing through that sanded nozzle you describe might breach that patent. Because you now deposite at that 90 degree angle, even though the rest of the printhead is at 45 degrees. But I'm not a lawyer, so I'm not sure.
I would be happy with taking the belt mechanism and putting it on a standard printer. Then add code that would push the finished part off and start a new part. I don't care about the long oversize prints.
Same. Seems like one step too far with the CR-30. People would love serial production off of a regular coreXY.
How about “squaring” the frame to make a Ender 5 like printer with unlimited Y? Not sure why it’s angled at all?
That's patented by MakerBot can't sell those.
Add 15 degrees to the head mount so that the nozel is vertical. it moves on xy plane still. Think that would need firmware change
be cool if the xy was horizontal to the belt, z was distance to the belt and "a" or "b" was the movement of the belt. so 4 axis.
I would rather see a traditional core XY with a belt as the bed where the XY axis move up and down and the belt doesnt move until you finish the print
I think the print mill is a good idea, and I'm sure over time support for it will become better, but I don't see how this is "better" than automating a standard printer at this moment. You don't need special tooling (special slicer) and the automation is almost as easy as just adding new gcode.
The only thing I can see is the ability to have a machine crank out a run of parts without you needing to be there. I often would like to set a second print that would finish before I get back to the office, but someone needs to remove the bed and set it for the next part.
Theoretically, the should be an ideal angle to mount the hotend at for reliability. It'll be between pressing the filament down properly and allowing clearance for the side of the hotend while things print.
Just had a thought, what if you file the top of the nozzle at a slight angle? I recently had trouble with a nozzle I enlarged being slightly off-centre causing bulges on one side (roughly equivalent to tilting the hotend at the contract point)
is there any reason why you couldn't put a 15* bracket on the hotend to set the hot end vertical? not familiar with this sort of printer
That wouldn't change the way the part is printed, but instead just create new issues since now your nozzle isn't perpendicular to your layer direction anymore.
@@MadeWithLayers thanks for the reply! I realised last night that 15* wouldn't work (i started with 45/45).
So for the 15* i think you'd want 75*, but the idea was that the nozzle would end up pointing directly down? Or is it already doing so and I'm missing something...
Id like a standard Cartesian printer with a belt bed for batch production of pieces.
Maker it has an patent on that until 2030 and they want crazy money for it.
@@FreeOfFantasy oh really!? I did not know that. Thats a pitty. Maybe ill get myself a printmill some day to mod into a standard style Cartesian. 😂
Maybe I missed it in the video but why is it even angled? Couldn't they just make it a normal design similar to the ender 3 with z rods and straight up and down but with a belt bed instead?
That's patented by MakerBot can't sell those.
What do you think about a 0 degree belt printer? Something that prints like a normal corexy but can eject the print afterwards?
From what I gathered there is a patent covering just that, which is why they use a 45 degree angle (which isn't covered in the patent). This means that if you want to sell that design commercially, you have to get a license from the patentholder, who can choose not to sell you one. Or can charge a hefty amount of money for that license, making your printer more expensive than theirs.
That was my original modding plan for the CR-30 😉
@@MadeWithLayers do you think your original plan would have been more of an improvement? 😉
I've been saying for a while that a belt printer could do a normal z layer build, but it would need more clearance behind the nozzle for the belt movement. We wouldn't get infinite sized objects, but we could get lots of small objects that had been designed for normal 3d printing.
Why not just G code a regular core XY head to knock prints off? This belt stuff is a total gimmick
@@ianhoolihan2396 Belt printers are a regular core XY head.
But I think what you're suggesting is to simply use a regular 3D printer movement system and have a way to knock the prints off. That's a great idea, and here's what I want you to do: Give it a shot and let me know how it goes.
There are lots of people who have given it a shot and have had various levels of success with it. However, it's not 100% reliable for all models. A belt is. "So why don't we put a belt on a normal movement system". Yup, Makerbot tried that. But it adds another layer of complexity. In simple terms, it adds another motor. Whereas, a tilted gantry like this gives you the advantage of a belt, with it's quick part removal and "infinite z", with no additional motors or complexity. However, it does change the game, and has it's own set of disadvantages because of that.
I wouldn't say belt printers are a gimmick. Belt printers are a new paradigm.
Question: why not just using a normal corexy, moving the belt and at the end moving the belt 1/3 of its lenght so you can make a Mass production which is fast like corexy and save Like the belt system 🤔🧐
At this point I'd like a belt upgrade for my Ender 5 pro. No infinite Z but automatic print ejection and otherwise the same machine with none of the angle issues.
Nice mod! But couldn't you just build a Voron 2.X style printsystem (with the printhead moving in core xy and z) over the belt to get a normal angle?
Yes, absolutely! But it can't be a finished product for the forseeable future since Makerbot has that patented.
@@MadeWithLayers Oh yeah, I remember the patent thing now. Bummer. Grüße aus Berlin!
What if you just completly modded it to a core XY style like a Hevo (hevORT) or voron with the belt moving up an down in the Z axis or XY gantree moving up and down for Z then the belt could also roll like normal giving it a infinite Y axis.
Then you wouldnt have to deal with this slicer issues of the angled gantree at all!
That would be a mod worth doing imo you would have the best of both worlds, core XY which has very good kinamatics, and still have a infinite axis in Y direction. It still would have software issues but atleased anything thats the Y axis travel for the print head will come out perfect every time. Printing in a infinite Y print like a 1.5m long cosplay sword would still run into slicer issues but your multiple small prints to the print head travel Y length will be all perfect.
I sooo want to see a time lapse of a flotilla of benchies being summoned.
They have no idea what I’m going to do with this, this is the start of my benchy army
Why don't they create one with a 90 degres angle like regular printer and replace the z screws with belts or some sort of counterweight to make it possible to make non planar move easily/quickly that would result with the option of printing at an angle for longer parts and as a standard angle for regular parts?
That's patented by MakerBot can't sell those.
Howany hours can you print continuously 🤔
I totally get how it looks like a Polaroid, once you said that I couldn't stop imagining a photograph coming out the front along the conveyor belt! 🤣 Thanks Tom, great video! 💜🧡💙
To push that idea a little farther, I wonder how well it would perform printing lithophanes?
congrats you almost make Cartesian printer
Wouldn't it be easier to just rotate the printing head (hotend) to be normal to be belt?
Wouldn't it just make more sense to have the printer in a normal config ie perpendicular, with the belt bed to enable the removal off prints ready for another print?
That's patented by MakerBot can't sell those.
Do we know why the belt is at an angle in the first place? Can you try angling the frame the other way so the belt is perpendicular to the other two axes like a normal X-Y-Z machine with the belt being one of those axes? My CR-10 and ender 3 call the axis which would be the belt "Y". So why can't the belt just replace the Y on an ender 3 or other conventional printer? It seems like all your problems are related to printing at an angle.
That's patented by MakerBot can't sell those.
why not 90° and have an infinite x or y and mount the z axis on something like a 't' joint in the middle or over the heated bed
What about bringing the angle from 45 to 60°?
Or indeed 90 degress, then lay it on its "side" and angle the nozzle according. Perhaps that would be a workaround for the patent issue.
I loved the montage music
I really don't get the angle thing, more calculations is more error prone. The idea of the bed is pretty nice however I use a magnetic rubber mat which helps to get it off easily.
Maybe someone needs to come up with caterpillar track conveyor with glass segments or fiberglass to improve adhesion.
Material pressure would want to push the part off the bed vs into the bed