A Slightly Different CoreXY Design

Hello All,

After recently learning about CoreXY, and having a fair bit of experience with both Cartesian and Delta printers, I had a design caught in my head that I just couldn't get rid of. Solidworks is a great antidote for that, and so I am presenting here some of my first attempts at turning my ideas into reality.

I wanted to incorporate linear rails and slides into this project, as I have had great experience with them in the past for projects like a 24x36 inch CNC router. I was surprised at the cost drop recently for smaller rails like the MGN12, so I decided to design the CoreXY mechanism around them. I also wanted to get away from the "cube of extrusion" style that dominates the CoreXY community - I have nothing against the t-slot rails and the designs that use them; but I just feel there are plenty of them around already, and I wanted to contribute something different.




Having graduated from PLA to ABS recently, I am becoming aware of the benefits of a fully enclosed printing space to maintain a more controlled ambient temperature environment. Instead of building an enclosure around the structure, I chose to incorporate the enclosure into the structure. Thus, the linear rails for Y and Z are mounted directly to side plates in a double "T" arrangement; ensuring that the rails are co-planar. I used 400mm rails for Y and Z, with a 450mm rail for X connected between them. I haven't yet determined the exact printing envelope, but it is definitely greater than the 300x300 I was shooting for. A back panel is incorporated currently as a 1/8" plexi part, but it might be beefed up to prevent racking of the sides, and a front cover is yet to be designed - I can envision a single or double hinged door, a sliding panel, or some other arrangement.



I tried to minimize the number of connecting parts, in part because I have always liked the idea of injection molding printed parts if volumes get to an appropriate scale. As such, I have reduced the number of unique plastic parts to 4: the XY Joiner/Z Support (6 instances), the Idler Support (2), the X Carriage (1 or 2 for dual extrusion), and the 30mm Fan Duct (2 per X Carriage). The Joiner is designed to be symmetrical so it can be used on both sides of the X rail, and does double duty by providing the connection between the Z Plate and the Z slides.




The Idler support is designed to mount on the same metal right-angle stepper mounts to guarantee pulley alignment,



and the X Carriage at the moment is a very simple Bowden mount with support for two fan ducts (which are probably overkill, particularly for ABS).




Variations on the X Carriage will be easy to derive, as they only need to implement the MGN12H hole pattern for mounting.

The side panels are interesting in that they are designed as a sandwich with a layer of 1/4" clear acrylic in the center (the outer panels at this point will probably be 1/4" MDF/Melamine, but I am also looking at foamed PVC, ABS sheet, and other materials). The idea is to place some of the LED lighting strip inside the inner panel to illuminate the side panel in an inset edge. I will probably start with RGB strip under Arduino control so that the color can change in various states, such as green for a completed print, or red (maybe flashing) to indicate errors. One benefit is that these color cues would be visible from across the room to indicate if attention is needed. Another thought is to use individually addressable NeoPixels to build a bar graph of print progress that would also be visible from afar.



I am completely aware that there is very little in the way of original thought in this design, but I wanted to present it here at this stage to get feedback from the community as to design deficiencies, things I might have overlooked, or to give those who might have developed similar designs an opportunity to contribute their experiences, good and bad. I tried to look through the archive for similar designs, but I didn't get too far (I am readily caught by a good build log). So if anyone knows of a similar design that they could point me to for study, I would appreciate it. And I am very willing to accept criticism of the design from those who know this field better than me - I would prefer to know the weak points now rather than having to redesign later.

I am slowly collecting parts to begin actually building a prototype of this design, and will post progress pics and renders as they develop. I am also willing to share my 2013 Solidworks files (and generic derivatives) for others to examine and modify once I am satisfied that they are worthwhile.

Thanks for your consideration and comment,

Dave

the printer is compact, well designed and is great looking, your design is great!
if it was me i would add a metal back panel to increase the ridigity and remove any possible lateral movement.
not sure if is a good idea to build this with mdf panel, way not metal plates?
If possible add the possibility to use direct extruder.
how do you add tension on the belts?
the 2 pieces that connect the y rails to the x rails will be printer?
i will fallow your project.

Edited 3 time(s). Last edit at 09/12/2016 08:48AM by filipeCampos.

Pretty lighting schemes are nice, but to me, they're sort of low priority compared to the stuff needed to actually make the printer work.

How many i3 people have problems with the Z axis motors getting out of sync? How are you preventing the same issue in your design? Using 4 bearing blocks on the Z axis will guarantee binding if there's any twist/flex in the frame of the machine. How will multiple layers of different materials behave when you heat up the inside of the box to print ABS? How will the mechanisms screwed to the sides of the box behave when the box is doing whatever it does when it heats up?

How will you tension the belts?

3D printers usually look nice until you start including all the stuff that actually makes them work. Then they get ugly. If you don't plan for everything in advance, you'll start tacking things in wherever they can be fit and your lovely design will start to look like every other undesigned 3D printer out there. Where are the X, Y, and Z limit switches? Where is the top cover (if you're going to print ABS you'll need it) and where will the filament spool be located, and how will filament be fed to the hot-end? Where will the electronics and wiring be located? How are you going to access them for service? Is this printer going to be tethered to a computer or stand-alone? Where will the control panel go? Is the bed heated? If not what are you printing on and how will the ABS stick? Where will you get heat for the enclosure (ABS needs 45-50C)?

---

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

filipeCampos:

Thank you for your compliments and suggestions. I realize I will probably need something beefier than 1/8" acrylic for the back panel, but I will wait until I actually build a prototype to try out alternatives. I know from cabinet building that a flat panel of even modest thickness can dramatically strengthen and square a box structure, especially when fastened in multiple points (not just the corners). The 20x20 extrusions serving as the rear stretchers will assist in that. Metal (probably aluminum) is on the table as far as panels go, but I worry about expansion under heating conditions. Wood composites like MDF might actually perform better. I plan on empirically testing materials in the prototype by mounting a dial indicator to a fixed spot (probably panel center) and indicating off critical surfaces like the Z rails as the interior is taken to operating temperature and back. I might do time lapse video with the indicator and a digital thermometer to get reliable data on how different materials perform, because at this point I don't have the experience to know exactly how things will behave.

I do have plans to make an X Carriage variant for a direct extruder. I have actually had reasonably good experience with the Bowden on my delta (longer than 600mm), but I know lots of people hate them; and I want to get into printing flexible filaments. So I actually have a direct drive on hand - the Toranado (http://www.thingiverse.com/thing:1246951) that I will model up an X Carriage mount for in the next couple of days.

The parts in red in my renders are printed parts. I will share the STL's and the source files for them (Solidworks 2013 plus IGS/STEP for those without) once they stop changing.

As for adjusting belt tension, I used a technique on my delta that I came up with on my own, but since have found it very common with users of belt clamps with the "teardrop" shape - simply get the belt reasonably tight when folding it into the clamp, then grab the loose end with a pair of needle nose pliers and rotate slightly to pull the belt tight a tooth at a time. Here is a render if that is hard to visualize:



I used this when building my delta, and I have never had to adjust belt tension since. I love a screw-adjust tensioner as well as the next person, and if I have to I am sure I could design a fixed part/moving part feature into the X carriage, but I will save that for when it becomes apparent that it is needed.

Digital_Dentist:

Thank you as well for your comments. Your points are all valid and thought-worthy. I agree that a pretty machine is useless if it doesn't work well; but I am really tired of so many designs that look like science fair projects I want to incorporate wiring channels and mounting surfaces into the multi-layer sides as well as bottom and back panels. I expect all my electronics to live under the bottom panel, where they can be accessed by either laying the printer on its back or by rotating the bottom plate up into the build chamber. I'll be integrating a control panel into the lower area as well. Thank you for reminding me about the limit switches - with a frame of extrusions these can be readily mounted as an after-thought, but I will need to place them appropriately. I will mount Y and Z on the side panels next to the rails, and probably place the X+ and X- inside the X carriage somehow, so that the wiring can fly with the extruder, temp sensor and fan wiring instead of having to follow along the Y.

I will be incorporating a heated bed - the renders show the over-the-top basalt glass 300x300 found here: [www.amazon.com] with probably this 120v 750 watt silicone heater: [www.amazon.com]

I'll likely have a cable chain carrying the wiring from the power supply and electronics following the Z plate. I'm thinking the top cover will be made from .060 Lexan, which readily cold-forms using a sheet metal brake.

I hadn't given any thought to providing heating for the build chamber. I just assumed the temperature would rise in an enclosed space due to the heated bed, . One thought that occurs to me is that Harbor Freight routinely puts their 1500 watt heat gun on sale for $8.99 - maybe the guts could be released and repackaged, and used with an SSR under software control to actively maintain a desired ambient temperature of the build space (does contemporary firmware support this?)

Thanks all again for the thoughts. I will post more once I incorporate these ideas into my design.

Dave

The only two advantages of the basalt are that it is flat and won't warp. However, it isn't thermally conductive and will take a long time to heat up with uneven temperature distribution (if you attach the heater to it directly). You'll need to set it on something to allow it to be leveled because you can't drill it very easily, and attaching a thermistor to it could be difficult. A piece of cast aluminum tooling plate is also flat and won't warp, but is thermally conductive and will heat evenly, probably a lot quicker than basalt. The price for the basalt is about the same as the retail price for a piece of 1/4" cast tooling plate that is the same size. You can easily cut and drill aluminum if needed. Aluminum won't break if you drop it or drop a tool on it.

The 750W heater is a bit much, but it's better to be overpowered than underpowered. If the machine is fully enclosed it may get warm enough to print ABS just from the bed heater. My current printer has a much larger enclosed space and the bed heater is sufficient during warm weather months. In the winter I have to replace a couple of the clear side panels with foam insulation board to maintain the chamber temperature.

I am working on a CoreXY design, too. I plan to use two lead screws on the sides near the front with companion linear guides in a more or less typical Z axis cantilever type arrangement, but with a third screw at the back of the bed to prevent it from bouncing like a standard cantilever. All three screws will be driven by one motor to ensure they remain synchronized. I will probably use those screws to level the bed. Once set, it should never have to be leveled again.

A 1500W heat gun would be huge overkill to warm the enclosure. A couple hundred watts should be sufficient. I installed a 400W heater (from a stripped Stratasys printer) in a Solidoodle printer and it heats the enclosure to 50C in about 5 minutes- about as fast as the bed heater gets the bed up to 105C for printing ABS. I used an external controller, but you can use an unused thermistor input on the printer's controller board and one of the MOSFET outputs to drive an SSR to regulate the enclosure temperature. You just tell the firmware that the enclosure heater is a second extruder. You can even run PID on it. You might want to install a fan to distribute the warm air inside the enclosure. I used a 220V fan, wired parallel to the heating element, running on 117VAC so it turns slowly and quietly, without blowing so hard that it screws up the print. You can buy small room heaters at your local Goodwill or SVdP store for $5 or less, but they'll disappear soon as winter is coming. Coffee makers, available from the same shops for $1-2, usually have a 1 kW heater...

If you put electronics on the bottom, I'd suggest using a removable panel so you can access the controller while the printer is upright and running (instead of laying the machine on its back), especially if the controller uses those awful little plug-in motor driver modules. It can be very useful to access the controller while setting up and troubleshooting the machine while it's running. I put the electronics in a drawer in my current machine and may do the same in the CoreXY.

---

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

I've tried two enclosures one just a simple insulated box around and old i3 which looked like a load of junk and only just worked, but I did manage to make a recirculating heated filtered air system for it with some tubing and a hairdryer to reference your heat gun idea (around 400w). The idea was to take the hot air at the top of the chamber and return it to the bottom hepa/charcoal filtered. Not so important with corexy as the heat is more useful at the top anyway.

I have thought about developing this into an external box which could be hidden under the bench with in and out hoses incorporating the heater and filters for all the functionality but none of the ugliness.

The second is a core xy with an acrylic enclosure. No motors or electronics in the chamber. It heats up without additional heaters to 50-60 deg C and I can switch on an 80mm fan to ventilate it if it's getting too hot and to (gently) speed up cooling after printing. Yes it looks a bit like a science fayre experiment but then it prints large abs items with ease. Functionally there are few compromises.



I do like clean neat machines too



But it's easier to package with a delta.

Edited 2 time(s). Last edit at 10/22/2016 08:28AM by DjDemonD.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Quote
devers6


and the X Carriage at the moment is a very simple Bowden mount with support for two fan ducts (which are probably overkill, particularly for ABS).

[attachment 83715 XCarriage.jpg]

Dave

Thats great look carriage but it is useless for ABS.(it is a design that would alow the easy addition of an led laser engraver.
You don't want air blowing down onto the print as it will cause the plastic to warp and separate, but you do want it blowing on the cold fins to keep the abs cool.

Edited 2 time(s). Last edit at 11/05/2016 09:53AM by GRAYWOLF.