Development corexy

I am developing a corexy having spent years working on delta printers its time for a change. This is work in progress as i develop the design in cad and publish its progress.

Specifications:

- build size 350x 350 x350
- heated bed with interchangeable build plates
- auto levelling using 3 z axis and probably a mechanical break z probe
- initially 1 hotend, direct drive then develop 2 hotends and then tool change system
- z retract on hotends
- electronics duet wifi with duet extension and paneldue
- fully enclosed with temperature controlled environment and extraction/filtered
- different printing medias

Picture of the progress below, so many features to design and finalise such as pulley/belt routing, hotend carriage, bed attachment to 3 axis plus many more features.




I look forward to any comments

---

www.andornot.co.uk, selling new products, including smoothie boards and other useful parts for 3D Printers:

AndOrNot buildplate launched, available in 6 inch, 8 inch, 12 inch and 17 inch square introductory discount code "20reprap"
And Or Not Build Plates

I like it

I'd put the steppers in the front and let the belts cross (or stack?) in the back over the centered Z screw.
Though keeping em in back helps tidy up wire routing.
Looking at your pic, If all 3 Z screws are anchored, makes me wonder if you need the Z rods at all?

Interesting concept to put the z-lift in the hotend. Also using non-captive leadscrew steppers is new to me.

One thing I'd reconsider is the placement of the linear rails. I'd try to put them below the top-members to reduce the gap between bed and X-gantry. Especially when using larger extrusion cross section. I'd also rotate the X-linear rail 90°. A vertical mounting plate makes it easier to feed filament from above.

When you say tool change system, do you think of using one direct drive stepper motor and only exchange the filament guide / idler gear ? That system wouldn't require one stepper for each tool and oozing of the unused tool is non existant. But you'd have to solve the problem of controlling the idler gear tension. Maybe use an electromagnet?

Edited 2 time(s). Last edit at 09/24/2017 02:30AM by o_lampe.

Quote
prot0typ1cal
I like it

I'd put the steppers in the front and let the belts cross (or stack?) in the back over the centered Z screw.
Though keeping em in back helps tidy up wire routing.
Looking at your pic, If all 3 Z screws are anchored, makes me wonder if you need the Z rods at all?

Thanks for your feed back

The 3 z screws are located in bearings top and bottom (not clear on the picture), allowing the linear screw to rotate, the stepper is non captive and will be fastened to a linear slide to stop it rotating (not modelled yet, i will post a picture later this evening). Then the stepper drives up and down the linear screw and will be attached to the bed with ball joint allowing the each of the z screws to level the bed using the routine available on the Duet firmware. The bed will be made from tooling plate, whether this will be heated or a plate above, i will know better after its been tested in our workshop.

---

www.andornot.co.uk, selling new products, including smoothie boards and other useful parts for 3D Printers:

AndOrNot buildplate launched, available in 6 inch, 8 inch, 12 inch and 17 inch square introductory discount code "20reprap"
And Or Not Build Plates

Quote
o_lampe
Interesting concept to put the z-lift in the hotend. Also using non-captive leadscrew steppers is new to me.

One thing I'd reconsider is the placement of the linear rails. I'd try to put them below the top-members to reduce the gap between bed and X-gantry. Especially when using larger extrusion cross section. I'd also rotate the X-linear rail 90°. A vertical mounting plate makes it easier to feed filament from above.

When you say tool change system, do you think of using one direct drive stepper motor and only exchange the filament guide / idler gear ? That system wouldn't require one stepper for each tool and oozing of the unused tool is non existant. But you'd have to solve the problem of controlling the idler gear tension. Maybe use an electromagnet?

Thanks for the advice, the x linear rail cannot be rotated by 90 degrees as the the moments and forces are the strongest in its current orientation, look at the design and suggestions, maybe a could change the slideways.

The tool change system will probably for a complete head and changed above where the x and y move into position and maybe the tool lowers and raises.

Using the duet allows me to expand and have many more stepper drivers and the paneldue gives the professional touch of a touch screen.

Mounting the motors on the bed plate means their cables have to move.

Mounting the linear guides and A and B motors on top of the frame members means you won't be able to put a top cover on the frame itself- your enclosure is going to have to extend above the XY stage somehow. I can't tell from the image but the if the Y axis bearing blocks overhang the edges of the frame, you won't be able to simply attach the enclosure walls to the frame- you have to stand them away from the frame somehow. Attaching enclosure panels directly to the frame adds to the rigidity of the frame (but you probably don't really need the increased rigidity). Is the enclosure is going to be a larger cabinet that this mechanism goes into?

The Z axis will be much less prone to wobble if you use fully supported linear guides instead of end supported round rails. Using two of them at the front of the machine and just the screw at the back will provide all the stability and guidance needed.

Where will the electronics and ventilation/filtration stuff go?

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Quote
the_digital_dentist
Mounting the motors on the bed plate means their cables have to move.

Thanks for the feedback
Yes the motors will move and the cable carried in an energy chain by igus, received some samples last week, they are perfect for keeping the cables tidy.

Quote
the_digital_dentist
Mounting the linear guides and A and B motors on top of the frame members means you won't be able to put a top cover on the frame itself- your enclosure is going to have to extend above the XY stage somehow. I can't tell from the image but the if the Y axis bearing blocks overhang the edges of the frame, you won't be able to simply attach the enclosure walls to the frame- you have to stand them away from the frame somehow. Attaching enclosure panels directly to the frame adds to the rigidity of the frame (but you probably don't really need the increased rigidity). Is the enclosure is going to be a larger cabinet that this mechanism goes into?

I may need to move the motors inboard slightly to accommodate the top cover, i envisage the cabinet to be part of the frame work, maybe attached using magnets.

Quote
the_digital_dentist
The Z axis will be much less prone to wobble if you use fully supported linear guides instead of end supported round rails. Using two of them at the front of the machine and just the screw at the back will provide all the stability and guidance needed.

i am using a three screw system for bed levelling, the end supported round rails are only being used as torque arms to prevent the motors turning.

Quote
the_digital_dentist
Where will the electronics and ventilation/filtration stuff go?

Probably on the top or the back of the printer.

My limited experience of using leadscrews suggests that it is very hard to eliminate wobble. So my choice would be to use two vertical linear guides or possibly round rods to define the horizontal position of the bed, one at the mid point of the left and right sides. Then one end of each leadscrew can be left unconstrained.

However, from your drawing, you appear to have the leadscrews passing through the motors. Is that right? If so then I guess that's why you have them mounted on the bed. I've not come across motors like that.

Edited 1 time(s). Last edit at 09/24/2017 12:41PM by dc42.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Quote
dc42
My limited experience of using leadscrews suggests that it is very hard to eliminate wobble. So my choice would be to use two vertical linear guides or possibly round rods to define the horizontal position of the bed, one at the mid point of the left and right sides. Then one end of each leadscrew can be left unconstrained.

However, from your drawing, you appear to have the leadscrews passing through the motors. Is that right? If so then I guess that's why you have then mounted on the bed. I've not come across motors like that.

Yes z axis design use non captive lead screw steppers, so to make the motors move up and down on the lead screw i just need to eliminate the twist/torque of the motors by using the linear slide. What i will do is model it up complete to tonight and post the design.

---

www.andornot.co.uk, selling new products, including smoothie boards and other useful parts for 3D Printers:

AndOrNot buildplate launched, available in 6 inch, 8 inch, 12 inch and 17 inch square introductory discount code "20reprap"
And Or Not Build Plates

If you're going for totally enclosed, might suggest moving motors outside the enclosure.

The A and B steppers could be below, and have a shaft through the center hole of the extrusion to the top where you'd mount a pulley.

One of the premises behind coreXY is no moving motors...
Dunno about the hollow Z motors, sounds like something you want to try rather than being practical.
Again, mounting the Z steppers below remove them from the heated chamber and eliminates the need for cable chains, something worth considering.

Quote
prot0typ1cal
If you're going for totally enclosed, might suggest moving motors outside the enclosure.

The A and B steppers could be below, and have a shaft through the center hole of the extrusion to the top where you'd mount a pulley.

One of the premises behind coreXY is no moving motors...
Dunno about the hollow Z motors, sounds like something you want to try rather than being practical.
Again, mounting the Z steppers below remove them from the heated chamber and eliminates the need for cable chains, something worth considering.

Good point about the position of the steppers, certainty position them out of the heat controlled environment.

---

www.andornot.co.uk, selling new products, including smoothie boards and other useful parts for 3D Printers:

AndOrNot buildplate launched, available in 6 inch, 8 inch, 12 inch and 17 inch square introductory discount code "20reprap"
And Or Not Build Plates

You might also want to consider the possibility of a simple upgrade to dual X carriages, such as the CoreXYUV configuration. Or even dual X and dual Y carriages.

Edited 1 time(s). Last edit at 09/24/2017 03:39PM by dc42.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

I cannot see any reason / advantage for moving motors on Z axis?
and
Motors where screws go thru the core?

If you must
one Z motor and a belt driving three leadscrew nuts on pulleys?

Maybe Inchworm drive

[en.wikipedia.org]

but driving leadscrews from base probably better?

Digital Dentist Core XY

I'm currently in the final design stages of a CoreXY, with similar goals to yours. Watch out for how much space 2x direct drive extruders take up. You soon find that the build table is too big and you lose printing space... it looks like this may happen to you. As for the Z lift on the hotends themself, only one of them needs to do this and the other can be stationary. The secondary hot end (the moving one) will either be in the active or inactive position (above or below the primary nozzle) and the build table can move to allow printing with whatever nozzle is active. This simplifies things a bit.

Other questions I've found myself asking during the design process...

What is the footprint of the printer? what is the build size? what is an acceptable printer volume to build volume? If this is a desktop machine, there is no point building a feature overkill machine that only has a tiny build space. Industrial machines don't care about this and build huge monstrosities that only make 200mm objects.

Where will filament go? Inside or outside the printer? How will it be routed to the hotend? The minimum bend radius of most filaments is not great meaning a lot of wasted space above the hotends. Most printers are not enclosed on the top for this very reason.

Motors inside or out? electronics inside or out? Important considerations for a heated enclosure.

Safety features? if you use a duetwifi without a screen and not within close range of a wifi device you may find yourself scrambled in the event you need an emergency stop. Simple to design in.... Also consider thermal fuses and other simple safety devices.

Thanks for all the advice and info so far, i will take a look at my design and consider the best way forward.

Quote
the_digital_dentist
The Z axis will be much less prone to wobble if you use fully supported linear guides instead of end supported round rails. Using two of them at the front of the machine and just the screw at the back will provide all the stability and guidance needed.

i am using a three screw system for bed levelling, the end supported round rails are only being used as torque arms to prevent the motors turning.


Bolting the motors to the bed support plate is what will prevent the motor bodies from turning. That's not what guide rails do.

Guide rails are used to constrain lateral motion and allow motion along their axes. One rail will provide Z guidance, but will allow the bed to rotate around it in the XY plane. A second rail, parallel to the first, prevents that rotation. That's fully constrained motion, under ideal conditions. End supported rails flex in the real world, especially when unstraight lead screws start putting lateral forces on the bed support plate they are lifting. That causes Z-wobble in prints and is the bane of many poorly made machines that use two screws to lift the X axis in Z. Fully supported linear guides are much less prone to flexure than end supported round rails, and the bearings are usually better, too (no play). If the two rails and screws are placed on one end of the bed, that's a cantilevered arrangement that leaves the opposite end of the bed free to bounce when the machine shakes from movement of the XY mechanism. Adding a third screw at that end of the bed supports it and prevents the bouncing.

Edited 6 time(s). Last edit at 09/26/2017 08:24AM by the_digital_dentist.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Quote
the_digital_dentist

Quote
the_digital_dentist
The Z axis will be much less prone to wobble if you use fully supported linear guides instead of end supported round rails. Using two of them at the front of the machine and just the screw at the back will provide all the stability and guidance needed.

i am using a three screw system for bed levelling, the end supported round rails are only being used as torque arms to prevent the motors turning.


Bolting the motors to the bed support plate is what will prevent the motor bodies from turning. That's not what guide rails do.

Guide rails are used to constrain lateral motion and allow motion along their axes. One rail will provide Z guidance, but will allow the bed to rotate around it in the XY plane. A second rail, parallel to the first, prevents that rotation. That's fully constrained motion, under ideal conditions. End supported rails flex in the real world, especially when unstraight lead screws start putting lateral forces on the bed support plate they are lifting. That causes Z-wobble in prints and is the bane of many poorly made machines that use two screws to lift the X axis in Z. Fully supported linear guides are much less prone to flexure than end supported round rails, and the bearings are usually better, too (no play). If the two rails and screws are placed on one end of the bed, that's a cantilevered arrangement that leaves the opposite end of the bed free to bounce when the machine shakes from movement of the XY mechanism. Adding a third screw at that end of the bed supports it and prevents the bouncing.

I think there is a misunderstanding on how i propose the z axis will work, let me draw it properly in 3d cad and then i will post it on this thread.

---

www.andornot.co.uk, selling new products, including smoothie boards and other useful parts for 3D Printers:

AndOrNot buildplate launched, available in 6 inch, 8 inch, 12 inch and 17 inch square introductory discount code "20reprap"
And Or Not Build Plates

Quote
andornot


I think there is a misunderstanding on how i propose the z axis will work, let me draw it properly in 3d cad and then i will post it on this thread.

By the look of just your Z system, it is clear it is over constrained. Concepts you seem to misunderstand.
A proper CAD package will show you what is wrong when you assemble the parts.
Now, quite frankly, you should try to copy a well thought, based on a sound design, working machine.
The choice is simple as there are not that many due to natural selection (lucky here it only applies to the design, not the designer)

---

"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)

Quote
MKSA

Quote
andornot


I think there is a misunderstanding on how i propose the z axis will work, let me draw it properly in 3d cad and then i will post it on this thread.

By the look of just your Z system, it is clear it is over constrained. Concepts you seem to misunderstand.
A proper CAD package will show you what is wrong when you assemble the parts.
Now, quite frankly, you should try to copy a well thought, based on a sound design, working machine.
The choice is simple as there are not that many due to natural selection (lucky here it only applies to the design, not the designer)

The z axis is not over constrained, and it is not a concept i do not understand. I use solidworks to design 3d printers and use it all day as engineering consultant to Research Centre designing bespoke machinery from concept right through to production. I have over 30 years designing machinery and i welcome constructive remarks and improvements to my designs but not insults.

I designed a delta printer from scratch 4 years ago and spent years developing and improving it, they even have one where i currently work and thought it was time to design a corexy and get design help from the forum users.

Quite frankly there are many ways to move the z axis, maybe this is not the best, but innovation and advancements in technology are not acheived by doing it like everybody else.

You were the first to claim we misunderstand, isn't ?
OK, then I will wait for your proper drawing.

---

"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)

Quote
MKSA
You were the first to claim we misunderstand, isn't ?
OK, then I will wait for your proper drawing.

I have read several of your recent posts and all you have done is criticised their designs, what machines have you designed?

You asked for comments.
This is like a critical design review (perhaps you have never gone thru one)
People will attack your design --- to make it better
or perhaps to say been there done that --- doesn't work very well

If you are the customer -- paying for design -- then you get to say
do it this way.

If you are an expert Mechanical Engineer
and expert in 3D printer design and development
Why did you ask for comments?

Just submit to Nobel prize commitee!

You're doing great andornot.
As a Engineer/designer/machinist, the whole review phase helps with direction.
Sometime "this is cool" is the only reason you need when you're in charge. It's nice when it's you, as often outside this forum the opposite is true, with everyone swinging their dicks and trying to leave their thumb print if things go right.

Quote
andornot

Quote
MKSA
You were the first to claim we misunderstand, isn't ?
OK, then I will wait for your proper drawing.

I have read several of your recent posts and all you have done is criticised their designs, what machines have you designed?

I have praised a couple of designs because they were sound AND/OR ingenious AND/OR well made.
NOT my fault if the majority of "designs" fail to meet these points, therefore my "critics" ! Fact is, none completed (some just computer renderings !), the threads are dead, failed or will fail once made (or will fail to sell)

So, waiting for you to give more details about your Z three motor/leadscrews/guides set up.
You wrote: "The 3 z screws are located in bearings top and bottom (not clear on the picture), allowing the linear screw to rotate, the stepper is non captive and will be fastened to a linear slide to stop it rotating " is quite puzzling for ex.


Where could I use XOR ?

Edited 3 time(s). Last edit at 09/29/2017 02:15AM by MKSA.

---

"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)

I've seen non-captive leadscrew steppers before and understood your Z-concept right away, ( bragging )
but if I'd go this route, I'd integrate the torque arm for the steppers into the vertical extrusions instead of adding rounded, end supported rods.

Dude... We understand your concept... I don't understand the hype with the captive steppers, i don't think its the best way to go, to me you are overcomplicating things , for 350x350 you can even get by with 3 regular leadscrews coupled with a closed loop belt and supported by linear rails(don't do round rods plz)....Oh and here's a "novel" idea,if you are so sought after using independent motors for each screw, i think that then you wouldn't even have to have linear guides if you use ballscrews contrained at the top/bottom(similar to what prototypical said just ball screws instead of lead screws)

Edited 2 time(s). Last edit at 09/29/2017 03:06PM by lukaCX.

Here is the design of my z axis in more detail:



you can see from this picture the three z axis supporting the bed, each of the columns have a ball joint connected to the bed to allow it to tilt by moving the any of the z axis columns. The ball joint also has movement in the x or y plane (depending which column it is) to allow for the change in length between the centres of the ball joints.



The view above shows more detail of each z axis column. The stepper motor is a non captive lead screw, i.e the lead screw nut within the stepper motor rotates and moves the lead screw up and down. If we were to hold the lead screw and stop the stepper motor spinning on the on the lead screw (due to friction) then the stepper motor would move up and down the shaft. So on my design the stepper motor is stopped rotating by the linear bearing held on by the red plate and therefore moves the ball joint up and down the z axis. From this you can control them to drive up to a z probe and level the bed.

I hope this explains it better!

It seems to me that every time a Z motor is started, or stopped, it's going to put lateral force on the end-supported guide rail (and the screw) which will flex and shift the print bed laterally, maybe causing z-wobble artifacts in the prints. All the bearings on round rails I've ever seen have some play (intentionally, to allow for rails that aren't quite parallel). Starting the motor will cause it to climb or drop on the screw a little as the bearing tilts, tilting the mounting plate and creating a lag between the motor movement and the bed plate movement, probably worse going up than going down. I think you'd be better off using linear guides mounted on frame members or flat plates. There's no play in the bearing blocks and fully supported mounting will minimize or eliminate Z-wobble. OTOH, printers don't normally print while the Z axis is in motion, so maybe any wobble created will settle out before the nozzle starts extruding again. Still, if the mounting plates tilt a little with each movement, you might lose precision and lift-on-retract could cause the bed to come back to a slightly different Z level than it started.

Kudos for mounting on three points and using spherical head screws. If you set one in a conical hole, the opposite one in a conical slot, and the third just against the flat bottom of the plate, the bed will be able to expand when heated without causing it to lift or buckle because of lateral force on the leveling screws. That arrangement allows for the bed to tilt, too. You'll need to add springs to hold the bed plate firmly against the screw heads. See [drmrehorst.blogspot.com] for a more detailed explanation of how I did it in my corexy printer.

Your CAD model looks like you have put bearings on the ends of the screws, but if I understand what you're doing correctly, the screws aren't going to rotate, so bearings won't be needed.

Edited 2 time(s). Last edit at 09/29/2017 05:27PM by the_digital_dentist.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Quote
the_digital_dentist
.....
Your CAD model looks like you have put bearings on the ends of the screws, but if I understand what you're doing correctly, the screws aren't going to rotate, so bearings won't be needed.

Congratulation doc, you are the only one who spotted the obvious fault ! In fact, the screw WILL rotate at an unpredictable speed depending on nut friction and counter motor torque. Imagine the result !

As for the rest, indeed plenty of question marks.The three ball supports, wrong, and more complicated to make/align than the correct solution !

Edited 3 time(s). Last edit at 09/30/2017 12:30AM by MKSA.

---

"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)

Ahh

As each hollow motor has its own rod (anti-rotation), you can have ball joints to the bed. Allowing for some float during homing and auto-leveling...