My first CoreXY!



  • Hello handicraft friends!

    Yes, I am also one of those people who want to make their own 3D-printer/or have made their own. My project took about 1 year from the first idea to the almost finished device, which is now standing in front of me. There were also some breaks and restarts.

    A few key data about the printer:
    Aluminium frame, CoreXY, print volume 300x300x350mm, Duet Wifi Board, 2x360 Watt power supply (one for heating bed only), Dualprint (with one Y-Splitter), BL Touch bed sensor, filament sensors, two Bulldog extruders, E3D v6 HotEnd, 24 Volt board tension,…

    Here are a few pictures of my construction, looks in real on almost so!




    In the meantime, the device has been running very satisfactorily for a few weeks.
    But as a developer you are never really satisfied with what you have created.
    My biggest worry is about the software I've adapted for this printer.
    It's all working, but is it all right? that question tortures me!
    So my question, is there anyone who can take a look at my. g files?

    BerTiNax
    (Sorry for this english, it was translated with DeepL)



  • Wow this looks really nice! What software did you use to design the assembly file?



  • Very impressive! Do you photos of the real machine?

    Is that a connector box, behind the Y split?



  • Pretty looking model. I note that there are 4 vertical guides for the Z axis plus two lead screws, all of which are constrained both top and bottom with no obvious means of adjustment. How did you solve the problem of getting all those 6 rods perfectly aligned such that the Z axis doesn't bind?



  • Ian, what would be an optimal design? I'm making a similar machine and the z-axis is the same as this one.
    Haven't cutted the aluminum parts yet, so there is time for me to change it.

    Delta guy here… a moving bed is unknown territory 😛



  • @Martin_S:

    Ian, what would be an optimal design? I'm making a similar machine and the z-axis is the same as this one.
    Haven't cutted the aluminum parts yet, so there is time for me to change it.

    Delta guy here… a moving bed is unknown territory 😛

    IMO the optimum design is to support the bed with 3 screws. Say one at each front corner and one in the middle at the back forming a triangle. Having just two screws means that the bed can always "rock". With the OP's design, without the linear guides, the bed could tilt from front to back. So the linear guides are being employed to constrain movement in all 3 directions which isn't what they are designed for. Any tendency for the bed to tilt will be transmitted as a bending force on the linear guides which could cause them to bind.

    Two points define a line but three points define a plain so having the screws positioned as I have described means that, as long as the screws are in sync, the bed will always rise and fall in a vertical fashion without any tendency to rock or twist no matter what load is applied to the surface. The screws should only provide lift - not guidance. Therefore they should be fixed at the bottom and where the nut attaches to the bed, but free at the top. Otherwise, they could act as guides which is not their job and it'll be dificult to align them with the "proper" guides. So now we have a flat bed that will stay flat as it rises and falls (in the Z direction), all we need to do is constrain it from any movement in the X and Y directions (i,e twisting when viewed from above). IMO this is best achieved with just two linear guides placed at opposite corners - say front right and rear left or vica versa. This will keep it rock solid and two guides are very much easier to aligned than four (or in the OP's case six if you include the constrained screws). The further the guides are away from any centre line, the more effective they will be. What I mean by that is that if there is say 0.01mm "play" in the guide and it's close to a centre line, that amount of play will be magnified considerably at the edge of the bed. Whereas, if the guide is placed at the edge of the bed, 0.01mm "play" is the worst case.

    The screws can be driven with a single motor and continuous belt which is how I've done mine, or with three separate motors. In either case, the screws can be used to level the bed negating the need for any other levelling screws, but there should be some flexibility in the screw/bed coupling to accommodate this. With three independently driven screws, one could use the Duet to automatically level the bed. With a single motor and continuous belt, levelling can be achieved by slackening the grub screw which holds the pulley to the shaft and rotating the screw. That's how I do mine but I've built my machine such that it stays level and so I've only ever levelled it twice - actually 3 times. The third time was when I dismantled the machine, put in my van and bumped it up the motorway to the TCT show at the NEC. The rear of the bed was low by about 0.5mm. Although after the return journey and reassembly at home, it didn't need any adjustment even though it had been disassembled again and bumped back down the motorway but I digress…...

    On the subject of screws, don't use those God awful 4 start 8mm lead screws - it's a pet hate of mine. See here https://somei3deas.wordpress.com/2017/01/31/z-axis-lead-screws/. If you are interested, you can see details of how I did my bed and the semi flexible bed/screw coupling here https://somei3deas.wordpress.com/my-corexy-printer-build/

    That's my twopence worth anyway. Others may disagree but it works for me.

    Ian



  • @deckingman:

    IMO the optimum design is to support the bed with 3 screws. Say one at each front corner and one in the middle at the back forming a triangle. Having just two screws means that the bed can always "rock". With the OP's design, without the linear guides, the bed could tilt from front to back. So the linear guides are being employed to constrain movement in all 3 directions which isn't what they are designed for. Any tendency for the bed to tilt will be transmitted as a bending force on the linear guides which could cause them to bind.

    Two points define a line but three points define a plain so having the screws positioned as I have described means that, as long as the screws are in sync, the bed will always rise and fall in a vertical fashion without any tendency to rock or twist no matter what load is applied to the surface. The screws should only provide lift - not guidance. Therefore they should be fixed at the bottom and where the nut attaches to the bed, but free at the top. Otherwise, they could act as guides which is not their job and it'll be dificult to align them with the "proper" guides. So now we have a flat bed that will stay flat as it rises and falls (in the Z direction), all we need to do is constrain it from any movement in the X and Y directions (i,e twisting when viewed from above). IMO this is best achieved with just two linear guides placed at opposite corners - say front right and rear left or vica versa. This will keep it rock solid and two guides are very much easier to aligned than four (or in the OP's case six if you include the constrained screws). The further the guides are away from any centre line, the more effective they will be. What I mean by that is that if there is say 0.01mm "play" in the guide and it's close to a centre line, that amount of play will be magnified considerably at the edge of the bed. Whereas, if the guide is placed at the edge of the bed, 0.01mm "play" is the worst case.

    The screws can be driven with a single motor and continuous belt which is how I've done mine, or with three separate motors. In either case, the screws can be used to level the bed negating the need for any other levelling screws, but there should be some flexibility in the screw/bed coupling to accommodate this. With three independently driven screws, one could use the Duet to automatically level the bed. With a single motor and continuous belt, levelling can be achieved by slackening the grub screw which holds the pulley to the shaft and rotating the screw. That's how I do mine but I've built my machine such that it stays level and so I've only ever levelled it twice - actually 3 times. The third time was when I dismantled the machine, put in my van and bumped it up the motorway to the TCT show at the NEC. The rear of the bed was low by about 0.5mm. Although after the return journey and reassembly at home, it didn't need any adjustment even though it had been disassembled again and bumped back down the motorway but I digress…...

    On the subject of screws, don't use those God awful 4 start 8mm lead screws - it's a pet hate of mine. See here https://somei3deas.wordpress.com/2017/01/31/z-axis-lead-screws/. If you are interested, you can see details of how I did my bed and the semi flexible bed/screw coupling here https://somei3deas.wordpress.com/my-corexy-printer-build/

    That's my twopence worth anyway. Others may disagree but it works for me.

    Ian

    Wow, nice. Thank you Ian for taking the time to do a detailed explanation, I really appreciate it.
    I think I understand what you mean, but there might be an issue with the stuff I already have in stock. I've 2 nema17 with an incorporated leadscrew (they are TR8*4). The plan was to get them in sync with dual endstops and do the autolevel with a z probe. So… I can't use a single motor to lift the bed. Adding another motor w/leadscrew would be a pita to sync right?

    So what I could do is use 2 leadscrews at each front corner and 2 linear guides in the back? I'm just guessing here.
    Or just do what you suggested NOT to do and make sure everything is perfectly aligned?

    I've other doubts about the bed design, but I'm going to continue in another thread.
    Sorry BerTiNax for hijacking your topic.



  • and yes… BerTiNax, post some real pics please! 🙂



  • Great models, looks like it will be great machine. Ian's advice seems very sensible.

    One question, since I have several enclosed machines, and your's seems to be enclosed, how do you intend to enclose the top?



  • @Martin_S:

    …................Adding another motor w/leadscrew would be a pita to sync right?

    No, you can drive the bed with 3 screws - Duet will take care of it and you can then use the auto levelling feature.

    So what I could do is use 2 leadscrews at each front corner and 2 linear guides in the back? I'm just guessing here.

    That just ain't gonna work.

    Or just do what you suggested NOT to do and make sure everything is perfectly aligned?

    One can lead a horse to water but one can't make it drink. That is to say, I've advised what I think you should do but it's up to you.

    I've other doubts about the bed design, but I'm going to continue in another thread.
    Sorry BerTiNax for hijacking your topic.



  • @DjDemonD:

    Great models, looks like it will be great machine. Ian's advice seems very sensible.

    One question, since I have several enclosed machines, and your's seems to be enclosed, how do you intend to enclose the top?

    According the OP it's built and running - shame we don't have any real pictures. I'd like to know about the thing that has an LCD which looks like it might be a PanelDue but with some other switches and a knob. It looks like it's inside the enclosure?



  • Hello handicraft friends!

    First of all, thank you for your praise of my work.
    I designed the printer with Autodesk Inventor because I also work with it professionally.
    As the name of this thread already says, it is my first 3D printer that I developed and built. Thanks to all of you who post their impressions and suggestions for improvement and thus help me with the further development.
    With regard to the building platform and the Z-axis, I would like to say that the threaded spindle is not clamped at the top, but only guided. Below at the engine is the fixed bearing, at the top is the floating bearing.
    The four guide shafts are adjustable by the brackets, which have a little play before tightening the screws. During assembly, the building platform is at the bottom first, then tighten the bolts of the shaft mount, then move the building platform upwards (without threaded spindle) and then tighten the upper screws. Only then are the threaded spindles and nuts assembled and screwed on in the same way.
    This really works very well. The mounting of the heating platform with only three screws is possible and a good choice. But I thought that four screws are a bit more stable and not statically overdetermined by the springs.

    The panelling with plexiglass has been ordered but not yet delivered. The upper end of the printer is not yet clear, but I have some good ideas.

    I didn't build the display in the 3D views. The DuetWifi can be used without a display.
    If you have any further questions or suggestions, please do not hesitate to contact me.

    BerTiNax
    (Sorry for this English, it was translated with DeepL)

    PS:
    Actually, I'm looking for someone who can view and check my configuration files (. g).









  • Super nice, really good job on both design and fabrication.



  • Yes, that looks really good and I'd say from that print that there can't be much wrong with your configuration files. Post them anyway and we'll take a look.



  • @Martin_S:

    Ian, what would be an optimal design? I'm making a similar machine and the z-axis is the same as this one.
    Haven't cutted the aluminum parts yet, so there is time for me to change it.

    Delta guy here… a moving bed is unknown territory 😛

    The OP has a great design. I use the same concept with mine, so does Stratasys with the Fortus series. This printer looks well designed and quite tidy.



  • Your printer looks really professional! Great job.



  • @rs4race:

    @Martin_S:

    Ian, what would be an optimal design? I'm making a similar machine and the z-axis is the same as this one.
    Haven't cutted the aluminum parts yet, so there is time for me to change it.

    Delta guy here… a moving bed is unknown territory 😛

    The OP has a great design. I use the same concept with mine, so does Stratasys with the Fortus series. This printer looks well designed and quite tidy.

    Fair comment but I've been helping a couple of guys from a company that I met while at the TCT show. They had problems with the Z axis binding because they couldn't get all 4 vertical rods perfectly aligned. So IMO it's not the best design concept.



  • I see that you have your belts in the same plane; I use the same config. Could you make a closer picture of the part where they cross?

    Thanks.



  • Hello handicraft friends!

    This is a small bridge that prevents the two belts from touching and rubbing against each other. It is a very simple and effective solution. (If you want, I can send you the. stl file.)




  • I have as similar system, with a PTFE sheet between belts. I just wanted to compare, as yours is smaller.

    Thanks for the picture!


 

Looks like your connection to Duet3D was lost, please wait while we try to reconnect.