I can see so many "why" in the comments but not a single "why not". Sometimes the journey is more important than the outcome. Let's go, new printer design!
Finally!! some new innovations! it seems to be incredibly hard to find videos about new innovations in the diy/3d print world. The videos are usually just the same 3 concepts being beat to death. But not you sir, you actually have a novel idea
@@DRedGuia you have to have a start with something. also your argument was used for traditional printers too "the z screw method is to wobbly, it will never happen" and look where we are now.
@@elementzero_0"start with something" means start with something that wasn't disproven already. This is a cool hobby project, but would absolutely not result in a good printer.
First, I love that you are experimenting with something different, that is very cool. I love that you have made it to the point that you have it printing. Also, very cool. One thing that does concern me a bit is that the rails are attached with 3D printed parts. I suspect over time these will want to sag and you will fight to get good first layers. Once you are happy with the design, I would want to spend the money to have these either 3D printed in Aluminum or CNC machined. Maybe even a bent steel bracket would work. Anyway, great job on this. It is nice to see something other than more CoreXY cubes. Cross gantry printers can be just as fast if done right.
Super cool. An auto bed levelling sensor would go a long ways towards compensating for any sag in the attachment. Could even fix the problem outright (if it even presents itself). I love the idea of having a fairly large printer this way, and being able to use the space as a desk or something when you're not printing.
please keep going with this, i hope that some cad files can be released soon. have you considered using 2010 aluminum extrusions to maybe stiffen up without losing too much volume? that could help with sagging etc
i think the overall design of this machine is awesome. it also can bring very good results to printing as the bed doesnt move in any direction. maybe there is a way to conter the weight of sagging by adding a weight on the other side?
I like cantilever printers, this is a nice way to get some stability at the far end of the rails. My 300x300x330 (cartesian) has independent steppers and drives on the z axis, so i only run z-tilt before a print and load my bed map. My cantilever printers just load a map and my first layers are better on the cantilever. Once leveled correctly, they seldom move. My 300 cubed, even with z-tilt set to .05mm, has a hard time getting exactly back to square after the steppers power down, and it shows on big prints. The cantilevers are mechanically squared and only move if impacted or reassembled. The only problem i have with them is that ringing is worse the further away from the mast you get, input shaper can only do so much, this crossed design should help to eliminate that!
I wonder if this design could have advantages over Core XY or bed slingers with regard to speed potential. Any idea how the head weight compares to the other kinds of machines? I also think that some of the consistency issues could be tuned in software once the sag/droop of the rails is modeled.
That's on my agenda, but there are some more important todos first. At the moment it is running Arduino Mega + RAMPS 1.6. If I'm not mistaken it is still possible to load klipper's firmware and install klipper itself on a linux notebook.
Seems like the further away from the main supported corner it gets a little worse. As much as this looks cool its going to be tough dealing with your toolhead weighted so far away from the main structurally supported corner.
@@CasualCNC Im not familiar with the rest of your setup but I know in my custom build all my 2020 extrusions needed to be measured with digital caliper and made sure everything was as little tolerances as possible otherwise even with my auto bed leveling if it wasnt perfect you would get frustrating results. Especially when it came to accuracy of print sizes.
@@AWSW-de But you have more linear rails and more linear bearings. it is more complex and high-quality linear rails and linear bearings are very expensive so it's also a lot more expensive. You could say that the belt system is simpler but you could use it on a standard coreXY frame. It's cool looking but I don't see any benefits.
Cantilever only deals with one z stepper and the sag from gravity. My 300x300x330 uses dual z steppers on their own drivers. Z-tilt set to .05. Gantry tilts every time the steppers unlock. Have to run z-tilt before every print and it still has issues with first layer on large prints. On a cantilever, leveling is all mechanical and you don't deal with xy tilting out of plane when the steppers unlock. The fault with cantilever is the further away from the mast you get, the worse ringing and positional accuracy. This design would help that a lot while not dealing with the issues of a dual z setup.
How smooth is it? It seems like the next step would be a decent quality CNC aluminum block to hold the rails onto the linear blocks. You’re probably getting some pretty bad deflection. Although, you may encounter some binding if you go with a more rigid carrier block. I can see that the mechanical will be difficult to tune to get good repeatability. Hope you can overcome the challenges. I can already see the main use case, which is a high density print farm. This things take up space and if you’re in a city space is money. Also, don’t mind the idiots. Idiots will idiot.
It is not rigid. Every single part of it. On one side, I'm surprised how good it prints and even at the max speed (before steps start being skipped) it does not fall apart, but the print quality is not sufficient. More calibration or input shaping would definitely improve it, but first I plan to take care of some low-hanging fruits to improve stiffness, like the main block, which also bends significantly. The density is not as good as it may look: The bed is 23.5x23.5 cm, but only approx. 16x16 may be printed, and the printer itself is 32x32.
@@CasualCNCthe density is actually already on par with bed slingers. I think you can beat the density within a couple more iterations. I have also always been fascinated with high density printing myself. CoreXY is decently dense, but the motion system is complex to tune and maintain if it ever goes out of wack. I would have never thought about a double cross cantilever. I was thinking more of a single moving Z gantry and single cantilever. Regardless seems like a really fun experiment.
Месяц назад
es genial, pero el peso del cabezal hace oscilar los extremos si se pudiera solucionar eso seria perfecta
@@DRedGuia actually it's not. Cantilever only deals with one z stepper and the sag from gravity. My 300x300x330 uses dual z steppers on their own drivers. Z-tilt set to .05. Gantry tilts every time the steppers unlock. Have to run z-tilt before every print and it still has issues with first layer on large prints. On a cantilever, leveling is all mechanical and you don't deal with xy tilting out of plane when the steppers unlock. The fault with cantilever is the further away from the mast you get, the worse ringing and positional accuracy. This design would help that a lot while not dealing with the issues of a dual z setup.
It would but it is at the bear minimum - it might seem bulky, but it contains rail sliders and has some empty space to accumulate XY-axis holders when moved to the far corner. And... actually from the print it is seen that the closest corner (visually) is being printed too high - there seems to be overcompensation for some reason.
I’m going to ask not because I want you to change anything but why did you choose marlin over kipper? I personally would have chosen klipper but I’m just curious to know your reasons for not.
First, I had some experience with Marlin when starting. Second, I have no dedicated klipper HW nor Raspi laying around. It runs on Arduino Mega + RAMPS so I would need to setup klipper firmware on Arduino and kipper server on a linux notebook. Klipper is definitely on my agenda later.
I think, this is mostly a calibration issue and not the hardware. Any tips how to improve? BTW, when printing ABS, I had to cope with the first layer much worse on K1C. This one sticks good and is actually quite usable.
@@CasualCNC It seems to be way too high in Z. And ABS without an enclosure might be a waste of time. Try PLA to get down good first, I think. The printer looks dope and I hope you get it working good. I would be interested in a kit. =)
@@AWSW-de By K1C I mean "Creality K1C" which has enclosure and prints very nice, but not big ABS parts due to bending of table between 50°C (bed leveling) and 105°C (printing temperature).
I've found the culprit: Print head was loose on the hotend. It is SO MUCH better now at 60 mm/s, the left corner is still not perfect - printed too high.
I can see so many "why" in the comments but not a single "why not". Sometimes the journey is more important than the outcome. Let's go, new printer design!
Why? Because it's freakin' cool!
@willcubemakes but absolutely pointless, it woud print at 500mm^2 max.
Any 200$ priner woud do 2000 to 3000
@@DRedGuia Speed isn't the only factor in a printer's usability
@@DRedGuia Since when do hotrods have to be practical?
@@DRedGuia "is slow therebefore is pointless"
Dude never saw a Prusa or Ender 3 💀
Finally!! some new innovations! it seems to be incredibly hard to find videos about new innovations in the diy/3d print world.
The videos are usually just the same 3 concepts being beat to death.
But not you sir, you actually have a novel idea
But absolutely pointless, it would be the worst printer, they just took the ultimaker mechanics and make it 10x worse
@@DRedGuia you have to have a start with something.
also your argument was used for traditional printers too "the z screw method is to wobbly, it will never happen" and look where we are now.
I know a new innovations. 😊
@@elementzero_0 Dude, there was no such printers for a reason
@@elementzero_0"start with something" means start with something that wasn't disproven already. This is a cool hobby project, but would absolutely not result in a good printer.
I love experimental printers
It’s official, I am in for the ride. Can’t wait to see more!
Absolutely love this, please keep it up
First, I love that you are experimenting with something different, that is very cool. I love that you have made it to the point that you have it printing. Also, very cool. One thing that does concern me a bit is that the rails are attached with 3D printed parts. I suspect over time these will want to sag and you will fight to get good first layers. Once you are happy with the design, I would want to spend the money to have these either 3D printed in Aluminum or CNC machined. Maybe even a bent steel bracket would work. Anyway, great job on this. It is nice to see something other than more CoreXY cubes. Cross gantry printers can be just as fast if done right.
Super cool. An auto bed levelling sensor would go a long ways towards compensating for any sag in the attachment. Could even fix the problem outright (if it even presents itself). I love the idea of having a fairly large printer this way, and being able to use the space as a desk or something when you're not printing.
Fantastic design, I just subbed for more interesting 3d Printer designs.
There was no printers like that for a reason
Bravo! Hope one day I will see your printer on the market :)
I know you probably already got a lot of comments saying to release cad files, i'm in too. this is a fantastic printer design.
please keep going with this, i hope that some cad files can be released soon. have you considered using 2010 aluminum extrusions to maybe stiffen up without losing too much volume? that could help with sagging etc
This printer is born dead already.
@@DRedGuia and? does that matter? always love seeing unique designs. good or bad it doesnt matter man, respect the effort bro
Keep going
i think the overall design of this machine is awesome. it also can bring very good results to printing as the bed doesnt move in any direction. maybe there is a way to conter the weight of sagging by adding a weight on the other side?
Hi, are you planning on releasing the cad files once finished? I would be interested
That might be an option. Noted.
I like cantilever printers, this is a nice way to get some stability at the far end of the rails. My 300x300x330 (cartesian) has independent steppers and drives on the z axis, so i only run z-tilt before a print and load my bed map. My cantilever printers just load a map and my first layers are better on the cantilever. Once leveled correctly, they seldom move. My 300 cubed, even with z-tilt set to .05mm, has a hard time getting exactly back to square after the steppers power down, and it shows on big prints. The cantilevers are mechanically squared and only move if impacted or reassembled. The only problem i have with them is that ringing is worse the further away from the mast you get, input shaper can only do so much, this crossed design should help to eliminate that!
I wonder if this design could have advantages over Core XY or bed slingers with regard to speed potential. Any idea how the head weight compares to the other kinds of machines?
I also think that some of the consistency issues could be tuned in software once the sag/droop of the rails is modeled.
This!
This is absolutely epic k would buy this
Some really cool stuff, would like to know how well it would work with input shaping when its printing
That's on my agenda, but there are some more important todos first. At the moment it is running Arduino Mega + RAMPS 1.6. If I'm not mistaken it is still possible to load klipper's firmware and install klipper itself on a linux notebook.
Are you using TMC Drivers?
In the meantime yes, TMC2209.
make the z axis retract in its entirety, this way u have a freaking small footprint
Seems like the further away from the main supported corner it gets a little worse. As much as this looks cool its going to be tough dealing with your toolhead weighted so far away from the main structurally supported corner.
Yes, but it seems to being overcompensated by the software for some reason: This corner is being printed with the head too high.
@@CasualCNC Sounds like the angle of your rails coming from your main corner are not parallel to your bed.
@@BrianVoelker This is very likely, but UBL probe and printing still use the same rails.
@@CasualCNC Im not familiar with the rest of your setup but I know in my custom build all my 2020 extrusions needed to be measured with digital caliper and made sure everything was as little tolerances as possible otherwise even with my auto bed leveling if it wasnt perfect you would get frustrating results. Especially when it came to accuracy of print sizes.
Very interesting! As a noob: what’s the benefits to a classic core xy printer? The far out corner isn’t supported here, isn’t that worse?
It is
But like how do you pull it off so easily
I'm struggling to see the benefits of this design, why would this be better than any other design?
Less parts? Maybe less complex, less price.. in the end you need to try it to know it
@@AWSW-de But you have more linear rails and more linear bearings. it is more complex and high-quality linear rails and linear bearings are very expensive so it's also a lot more expensive. You could say that the belt system is simpler but you could use it on a standard coreXY frame. It's cool looking but I don't see any benefits.
Cantilever only deals with one z stepper and the sag from gravity. My 300x300x330 uses dual z steppers on their own drivers. Z-tilt set to .05. Gantry tilts every time the steppers unlock. Have to run z-tilt before every print and it still has issues with first layer on large prints. On a cantilever, leveling is all mechanical and you don't deal with xy tilting out of plane when the steppers unlock. The fault with cantilever is the further away from the mast you get, the worse ringing and positional accuracy. This design would help that a lot while not dealing with the issues of a dual z setup.
How smooth is it? It seems like the next step would be a decent quality CNC aluminum block to hold the rails onto the linear blocks. You’re probably getting some pretty bad deflection. Although, you may encounter some binding if you go with a more rigid carrier block. I can see that the mechanical will be difficult to tune to get good repeatability. Hope you can overcome the challenges.
I can already see the main use case, which is a high density print farm. This things take up space and if you’re in a city space is money.
Also, don’t mind the idiots. Idiots will idiot.
It is not rigid. Every single part of it. On one side, I'm surprised how good it prints and even at the max speed (before steps start being skipped) it does not fall apart, but the print quality is not sufficient. More calibration or input shaping would definitely improve it, but first I plan to take care of some low-hanging fruits to improve stiffness, like the main block, which also bends significantly.
The density is not as good as it may look: The bed is 23.5x23.5 cm, but only approx. 16x16 may be printed, and the printer itself is 32x32.
@@CasualCNCthe density is actually already on par with bed slingers. I think you can beat the density within a couple more iterations. I have also always been fascinated with high density printing myself. CoreXY is decently dense, but the motion system is complex to tune and maintain if it ever goes out of wack. I would have never thought about a double cross cantilever. I was thinking more of a single moving Z gantry and single cantilever.
Regardless seems like a really fun experiment.
es genial, pero el peso del cabezal hace oscilar los extremos si se pudiera solucionar eso seria perfecta
I would love to see openscourced files for this cuz its so cool :)
You can actually hear the acceleration ramps. This has got to be like stock profile Ender 500 mm/s² range...
You are right: I've used Prusa Mini's profile as basis in Orca and it has acceleration set to 500 mm/s² for the first layer which I've overlooked.
@@CasualCNCyou know your design is pointless right?
@@DRedGuia actually it's not. Cantilever only deals with one z stepper and the sag from gravity. My 300x300x330 uses dual z steppers on their own drivers. Z-tilt set to .05. Gantry tilts every time the steppers unlock. Have to run z-tilt before every print and it still has issues with first layer on large prints. On a cantilever, leveling is all mechanical and you don't deal with xy tilting out of plane when the steppers unlock. The fault with cantilever is the further away from the mast you get, the worse ringing and positional accuracy. This design would help that a lot while not dealing with the issues of a dual z setup.
Hey cool it works
I feel like making the tool head lighter would improve print quality on the outer regions.
It would but it is at the bear minimum - it might seem bulky, but it contains rail sliders and has some empty space to accumulate XY-axis holders when moved to the far corner.
And... actually from the print it is seen that the closest corner (visually) is being printed too high - there seems to be overcompensation for some reason.
I’m going to ask not because I want you to change anything but why did you choose marlin over kipper? I personally would have chosen klipper but I’m just curious to know your reasons for not.
First, I had some experience with Marlin when starting. Second, I have no dedicated klipper HW nor Raspi laying around. It runs on Arduino Mega + RAMPS so I would need to setup klipper firmware on Arduino and kipper server on a linux notebook.
Klipper is definitely on my agenda later.
Nice!!
that's so perfect😂
whats the advantage? less parts because end dont have to be supported?
Nice Printer! Are the Files available?
Not at the moment, it is a work-in-progress.
How is the sack of the rails with printhead and corresponding axis at the furthest point?
The measured heights at the moment (G29 P1):
Front: +0.45 Back: +0.01
Left: +0.63 Right: -0.32
What advantages does this printer have then others?
That is really bad first layer, keep going and good luck ;)
I think, this is mostly a calibration issue and not the hardware. Any tips how to improve?
BTW, when printing ABS, I had to cope with the first layer much worse on K1C. This one sticks good and is actually quite usable.
@@CasualCNC It seems to be way too high in Z. And ABS without an enclosure might be a waste of time. Try PLA to get down good first, I think. The printer looks dope and I hope you get it working good. I would be interested in a kit. =)
@@AWSW-de By K1C I mean "Creality K1C" which has enclosure and prints very nice, but not big ABS parts due to bending of table between 50°C (bed leveling) and 105°C (printing temperature).
@@CasualCNC Hi, ah sorry, got that wrong. Thanks for clarifying =)
I've found the culprit: Print head was loose on the hotend. It is SO MUCH better now at 60 mm/s, the left corner is still not perfect - printed too high.
This printer was dead even before construction