Hacking the CR30 to print in layers

[@3dpprofessor]

You into Math? www.3dpprofessor.com/2021/11/10/hacking-the-cr30-to-print-in-layers-video/

[@williamsteele]

So, the one big problem with moving the belt backwards that I don't see you've accounted for is the flex in the belt. If you move it in the same direction, you're always putting the belt under tension... however, if you move it backwards you'll actually pull all the slack and put it onto the top of the belt (normally, it's pulled taught.) That difference can actually account for an error of several hundred microns... and unless you have a motor at each end, you won't be able to a) eliminate it or b) compensate for it as it'll vary depending on what's printed. In my original concept printer, I actually did try it.

[@3dpprofessor]

@@williamsteele Others have brought that up, and all I can say is "wasn't a problem in my tests". Yeah, it might be off by a little, and it might be worse for bigger things. But that's why we need to get this slicer mod, so we can test it!

[@NAK3DDesigns]

Love the video, Thanks for the shout out, & NEVER stop trying new things that's how we make this stuff better..

[@Audio_Simon]

I love the idea of a flip bracket on the head. Creality should include it as default if it doesn't get them sued.

[@dstewar]

That was my instant thought. One that just let you flip between vertical and oblique

[@samuel_hafen]

I think it would be possible to create a custom profile for a mainstream slicer. This profile can implement layer sift on each layer and remap the z axis, so that no modification at the piece you want to print is necessary. So you easily can batch print small parts. This would be ideal for me, i bought a CR-30 and did not realize that the parts which i want to print are impossible to print on it. My printer was just expensive junk upon this point.

[@3dpprofessor]

Hope this helps.

[@satibel]

there should be a relatively trivial way to make it only do downstrokes, keep track of the position, then for each move if the Y direction is negative, chain the movements in a list while the direction is negative, then add a hop and a movement that goes to the end of the last one, and reverse each movement, and add the same hop and movement after that. you will always get 2 extra movements after each chain though.

[@ddnguyen278]

3D printing tech is evolving so fast!

[@Leonardokite]

Now there's a man who knows his 3D printers! And when he has a good idea, he sticks to it like a dog on a bone.

[@PrintingNerd]

Looking at the patent you mentioned Joe could also apply to the CR30 or any other belt printer as the patent is specifically on belt printing and continuous fabrication. So the patent is for a belt print surface not the orientation of the part or if it's one printed continuous or multiple copies printed. The conveyer bed is the core if the patient

[@3dpprofessor]

Possibly. But the patent doesn't mention continuous printing. Just sequential. But, again, being sued is less about the patent and more about the disposition of the patent holder, and so far Stratasys has been strangely quiet.

[@PrintingNerd]

@@3dpprofessor Agreed. Which I'm surprised with as they have been relentless in the past.

[@ogpennywise]

Sounds like the belt printer bed modification for the Voron v2.4r2. I haven't seen one in action though.

[@kamehax]

quite impressive where can we find the bracket? i would love to try it out

[@claws61821]

While I applaud the mention of calligraphy nibs and the attempt to find a purely software solution, I personally think that a better solution - and one more popular due to preexisting desires for it throughout much of the community - would be too add a rotary axis either at the print head or at the mounting points for the x-axis gantry. This would also require new slicing profiles, but it's already sought-after, it has much greater versatility, and it avoids several of the complications of the novel "calligraphic 3d printing" concept you propose here. That IS a really cool idea though.

[@tasmedic]

Lovely video, thanks. Please check your audio. It seems to be clipping.

[@francisduvivier]

Interesting. As a quick hack in between, you could post modify the normal gcode to modify all movements going against the grain to - disable extrusion - add a small Y-hop - then do the opposite movement with extrusion enabled - then do the movement with hop again to end up in the correct place

[@3dpprofessor]

Yes, but how to identify the gcode movement that go against the grain, and what do you do when printing something, like a Chibimal, with a lot of small against the grain movements in sequence? I'd be interested to know how this can be done.

[@francisduvivier]

@@3dpprofessor If the movements are linear, identifying the gcode shouldn't be too hard. If they are not, not sure what to do then. And those very small pieces would indeed by quite a pain. But i think there's a doable way to group them in a lot of cases.

[@3dpprofessor]

@@francisduvivier I'll say what I often say in the makerspace: "Give it a shot and let me know how it goes." You don't need a belt printer to try and see if you can cancel any negative Y movement.

[@gcod3d161]

A way to group them is writing a post slice script to look ahead to the next gcode line and compare that Y move to the previous Y move. Assuming you want all moves to move in the positive infinite Y direction (or “with the grain”) of an XY coordinate plane, if the next coordinate has a Y component larger than the previous coordinate then you’re moving in the positively infinite Y direction, if it’s smaller then you’re moving toward the negatively infinite Y direction (or “against the grain”). If the Y value is the same in the next extrusion gcode line then that means it’s only an X move, it could be ignored or used to mark the end of a smaller group of consecutive moves. reverse all the against the grain moves to make them with the grain and get to printing. speed increase I can think of off the top of my head is making the infill angle at 90 degrees so every other layer is mostly made by X moves, probably would be best to start from the “negative” most coordinates so that you’re always moving with the grain, saving some time

[@3dpprofessor]

@@gcod3d161 Well, it's a simple matter of coding, then. Which, if you have ever worked in software, that statement has a reputation.

[@tacticalpoet]

They may have identified it as non enforceable due to prior art, as the concept was separately published on a 3d printer forum before the patent was filed.

[@sunekeller9547]

Could you link to the forum?

[@tacticalpoet]

@@sunekeller9547 that worked, sadly linking is hamstrung due to (fully justified) anti spam bot messages

[@tacticalpoet]

Or not.... its on the rep rap blogs 2008 November (11) idea-for-reduced-cartesian-structure

[@sunekeller9547]

@@tacticalpoet Thank you very much 👍

[@christiansrensen3810]

Your idea for down motion, is actually really needed for slicer on belt printer. I noticed higher thin Beams, are exposed to more then one issues. When the nossel travels upward, in lift the print, really just a little...(really minor). This affects the speed, you can go in the total print. When you slowmotion the corners in pushes just really tiny on the corners. This mean the line is just not get precision you want. So your idea for this, i would think could enhance speed quite a lot on the belt printers....check it out. Hollow beam. No more then 50mm wide but double the height 100-120mm

[@MrBaskins2010]

can you do a followup video better demonstrating the 90 degree print mod?

[@3dpprofessor]

I honestly haven't done much with it and I've taken the 90 degree mod off since the sicers don't yet support it.

[@doglarosa5418]

My first kit was a Thing-o-matic I had forgotten about the belt because it never worked right

[@3dpprofessor]

Didn't stop them from patenting it.

[@doglarosa5418]

@@3dpprofessor that is the funny thing about patents isn't it?

[@Gameboygenius]

Have you done any backlash testing on the belt axis? I would expect quite a bit of it because the belt might be a bit stretchy, as well as having friction.

[@3dpprofessor]

How would you test backslash? The prints turned out okay.

[@reasonablebeing5392]

I did not read through all of the comments but let's take your great idea to the max: replace the x-y gantry with a delta gantry - You would gain the speed of the Delta combined with the ability to angle the nozzle or make it orthogonal to the belt like a conventional 3D printer. That would be my ideal!

[@3dpprofessor]

That's an idea, but it loses the advantage of a belt printer, that belt printers do what they do with no additional complexity over a regular cartesian 3D printer. I mean, if you're going to use an extra motor so you can angle the print head, then why go full delta? Just put a motor on your print head, right?

[@reasonablebeing5392]

@@3dpprofessor In my mind you'd get the best of both worlds as it can function either way - classic or belt. Hopefully someone tries it and we can see the results.

[@Gunstick]

Only 10cm back move? And needs tilted head? Well do both: a second head on the gantry for layer printing. So use the extruder depending on the way you want to print.

[@jeffburns2179]

Where is the printer controller with the click wheel?

[@3dpprofessor]

Wherever you want it. I moved mine to the upright so I can reach it from the wall.

[@snotode543]

Could it be more efficient to tell the firmware that belt is y-axis and diagonal is z-axis*0.707? Then you dont need to modify gcode, not scewing the model in unknown best orientation

[@3dpprofessor]

That's not the way they've done it, but that could be another soltuion.

[@FunDumb]

Could this potentially be modified in the gcode. And just have chatGPT make the modifications?

[@3dpprofessor]

I'd say to give it a shot and let me know how it goes. However, my expectation based on past experience with ChatGPT is you get about 2 paragraphs that seem reasonable before it has a stroke and tells you to Gump the head at the front and tie Jack's feet together. I'm not sure I'd trust it with the operations of machinery just yet.

[@bobf3598]

I'm just glad I never purchased this printer. It might benefit from certain model designs, but this thing takes up so much real estate as well. As well as the nozzle being parallel to the bed which is what I didn't get when they designed this, but why not have have the z axis move in the vertical, like a cnc??

[@3dpprofessor]

"why not have have the z axis move in the vertical"? Probably because that would tread too closely to the patents held by makerbot.

[@williamsteele]

Curious... what about your part geometry is the CR-30 unable to print? Have you thought about printing them at a 45 degree angle with a corner pointed forward?

[@3dpprofessor]

Like at 3:24?

[@williamsteele]

@@3dpprofessor Um... yep... exactly. How did I miss that? :-D

[@williamsteele]

@@3dpprofessor One thing you can do on those overhangs to minimize them is to use a variable layer height and go for thin layers in just that section. (Or in any areas where you want a higher quality overhang on the front.)

[@Y3SS1N]

Why not print a 45 degree triangle first an then on that angle the print you wanna do.

[@3dpprofessor]

Besides the constant waste of a raft wedge, yeah, that would work.

[@Y3SS1N]

@@3dpprofessor sorry for my terrible English

[@grantbaxter554]

Great video as always, really nice idea (not one for how I use my machine, but I love it...). Couldn't you calculate a calibration cube and just manually create the G code as a POC? (as you said, not saying it's trivial, but it's a lot faster than trying to modifier a slicer...)

[@3dpprofessor]

I don't know. I'd say give it a shot and let me know how it goes.

[@grantbaxter554]

@@3dpprofessor Would love to, but I don't have a belt printer, only a delta

[@tmarti69]

I don’t know how absolute a patent is, they must enforce their patent, but they are not actually using their patent if they don’t produce, or license it. So, the gray area may indicate they have lost it already.

[@3dpprofessor]

That can turn around with one lawsuit. They can simply claim that they weren't aware of things before that point. Right now, pretending they don't see these belt printers, so they can claim the hobby market is too small for them to notice until it isn't, keeps their options open for a suit in the future. However, they'd be fighting internationally, there's no guarantee they'd be able to collect even if they win, so maybe they're letting Creality do their thing until they can find a domestic manufacturer they can actually collect from if they win. Who knows. The world of law is a weird one.

[@tmarti69]

@@3dpprofessor Blazing Saddles, do as little as possible to keep the patent, they haven’t done any evidence of that. What in current history have they done to progress on the patent technology they have been entrusted? I think there as an argument that if they do not license it out they have lost the patent. You could consult an actual patent attorney like Jones Waldo. They might be entitled to a piece of it though, but not all of it after all this time.

[@3dpprofessor]

@@tmarti69 I think you're right. But at the same time, that's for the actual courts to decide, and Stratasys has to decide if they want to take that risk.

[@tmarti69]

​@@3dpprofessor The intent of the law is to reward innovation; if none is being perused then there is no absolute patent.

[@DuelScreen]

It sounds like this would be trivial for Naomi Wu and her team. They have their own slicer code, the tech skills to modify it, and it would be in their financial interest to make their printer work both ways. Why don't you reach out to her team?

[@shaunmorrissey7313]

What? do some background research first before making stupid comments

[@DuelScreen]

@@shaunmorrissey7313 How is my comment stupid? What is it that you think you know that you think I don't?

[@3dpprofessor]

@Shaun, that was not appropriate. There's nothing wrong with @DuelScreen's comment. "Trivial" isn't the right word for what this is. It's gonna take a lot of work to implement. And Creality might not be keen to do it. Naomi has to push hard to get them to take a chance on the CR30 and the campaign didn't go as well as the leadership at Creality wanted it to, so the chances that they want to dump more time and resources into it, especially for something like this, is not high. That said, it's also not zero. Which is why I put this video out there. To clarify something that I have been trying to convey in text, prove the concept, and maybe inspire the development of something new.

[@wolkaiserdrake9946]

so, basically a normal printer but its bed is a conveyer

[@3dpprofessor]

And the tilted build gantry.

[@brunodherrera]

it's easier to just solve the "waste" recycling in an affordable way for once.

[@Actio83]

So annoying that a patent holds the entire printing community hostage…

[@3dpprofessor]

I mean, technically a patent held 3D printing hostage for a very long time. But that's the points of patents. And for as annoying as they are, looking back at history I still think they do more good than bad.

[@vasili1207]

Print times would skyrocket

[@3dpprofessor]

Where you gotta be in such a hurry?

[@Shrouded_reaper]

Printer slave works 24/7 though

[@niceau2010]

Calligraphy.. That's how we write in Chinese. The strokes are always from top to bottom, left to right.

[@3dpprofessor]

Yes, though as I understand it in Chinese the reason is that traditionally it was pained with a brush who's bristles get bent in a certain way, and so you have to go with that flow to keep the letters looking good. When I was in Singapore I learned how to count strokes in Chinese characters so I could look them up in my Chinese English dictionary. I never did get the order exactly rigth all the time, but it was enough to help me learn a little.

[@Lulzigi]

Did anybody ever figure out calligraphy slicing?

[@3dpprofessor]

I'm afraid no one is developing belt printing these days.

[@Lulzigi]

@@3dpprofessor that's unfortunate.

[@OldCurmudgeon3DP]

My question... The patent exists, but have they done anything with it in the last 7yrs? I'll admit I know nothing of their product line. However, if they haven't been producing devices (affordable or not) using the patent then shame on them.

[@3dpprofessor]

As far as I know, besides the belt plate for the cupcake, they never have done anything with it. But sadly, that doesn't invalidate their patent.

[@OldCurmudgeon3DP]

@@3dpprofessor true, but it makes them the "bad guys" for sitting on it. Stifles innovation in many respects. I believe if you're going to patent something then you should have plans to pursue development. Them appearing to have just sat on it for nearly a decade is wrong. Kinda hoping the kickstarter succeeds.

[@3dpprofessor]

@@OldCurmudgeon3DP that's kinda not the point of patents. Patents encourage innovation by giving inventors a safe place to share their ideas. If anyone "steals your idea", without compensating you, then the owner has rights to sue... until the patient expires. There problem is someone came up with a novel implementation and developed it before the patent expired.

[@OldCurmudgeon3DP]

@@3dpprofessor I get the purpose of patents. I don't believe they should be used for squatting on a concept as may be happening here. What I take away from your explanation is that a patent protects "an idea." If the patent holder does nothing with that idea for a decade, how does that benefit anyone. A useful patent allows the holder a chance to recoup R&D and turn a profit for a number of years; After which, others are free to run with the original design. Either pursue development & marketing, work w/ others to license the idea, or don't bother in the first place. 🤷🏼‍♂️ They already have a bad wrap over the initial patent that is said to have stifled printer development until it expired. Then again, I'm just a nobody and probably don't know what I'm talking about. (most likely)

[@claws61821]

@@3dpprofessor Actually, patent law *does* have requirements that the technology be used in production either directly or by licensed entities to remain valid after a specified time (I think three years). Judges working such cases with the USPO have a long history of reassigning patents or declaring them public material when someone else has infringed and the original patent holder hasn't used them in or since that period and cannot provide proof of imminent use already being spun up.

[@charbelbounader2404]

I really don't think the idea is worth the print time it's already long enough as is

[@ambujarajan]

So belt printers too problematic???

[@3dpprofessor]

Not _too_ problematic. It's a trade off. Every manufacturing process has it's limitations. Belt printing's limitaitons are just different than normal 3D prining.

[@neilmacgregor4680]

That's infinite-Y, not infinite-Z

[@3dpprofessor]

6:30 The way I had to edit the GCode disagrees with you. The printer think's the belt is Z, so it's Z. 0:56 And, if you stand up the printer like this, you can very quickly see how it could be a tilted Z. So it makes a lot of sense. However, this whole argument is muddied by the fact that we can't even agree which way is up, Z or Y, between different applications in the real world. Sometimes, even the same application will sometimes use "Z" for up, and sometimes use "Y" for up. Here's a classic 3DPProfessor video about this: kzbin.info/www/bejne/hqeyoGSLd659ntE

[@williamsteele]

I invented it... it's definetly the Z axis, which he described perfectly. Technically, that axis in my original code was Z' (Z Prime). The inputs to calculate that are Y and "desired" Z... which results in the Z' (belt distance.)

[@Kritkingpenguin]

Do you remember me

[@Gorilla_Jones]

I never see anything useful printed. Weird.

[@3dpprofessor]

Useful prints don't make interesting views.