I am designing a simple/cheap #CoreXY#3dprinter based on square tubular aluminum (steel?). The pieces need not be cut to perfect 90deg ends as they are not going to be connected end to end like most extrusion-based frames. They will be connected similar to the way scout towers are being constructed (more like the original quadrap-3d-printer to give a relevant example).
Because I don’t have a (T)slot to easily attach the #3dprinted components I have used a clamping mechanism for the square tube that is repeated throughout the design.
I have designed it in #DesignSpark Mechanical so I can only offer a #Sketchfab link for 3d viewing. 2 links actually, one the rendered STL …and one the Sketchup file I was able to export from DSM:
Here’s the more colorful #ScoutCoreXY sketchup import into Sketchfab (it misses the “duplicate” parts from the original file; for some reason sketchup only shows those once)
… and the more complete but missing color STL render (on Sketchfab too):
I’d like any opinions, etc (given the constraints of the design).
Very detailed model. I will mention from my past experiences with corexy a good method for adjustment of belt tension (and balancing tension of the two belts) is a key design consideration. Also how the belts force is applied near the top of the side rail carriages will allow them to act as a fulcrum which may end up translating to the hot end through the horizontal center rods… But that is just my theory. In practice it may not be an issue, but something to consider.
My last suggestion is on the bed. If you can try to allow for a three/two/one alignment. Three points to establish a plane, two points for a line, one for a point. So ideally you would have three cords defining the bed level, two vertical rods to lock it to a line, one point to snap it to its location on that line.
But overall I want to say great work. Please keep us updated on your progress. It reminds me a lot of the Quadrap frame. And I know many people have been very happy with that design by @shauki , and additional work from many great builders/modelers like @Michael_Memeteau . Keep up the great work.
@Eclsnowman
Thanks for your comment. I started designing a similar model as a repstrap - around the steel frame of a cube coffee table- but gave up that model because I wasn’t really happy without a more “standardized” design. This is a continuation of that model so yeah, it became quite detailed over time. It also has all the “lesson learned” from the failed parts and design mistakes of the previous model.
Regarding the belts tension, the Z is the only axis with an embedded tensioning mechanism. For XY I was looking at inline tensioners, those that you snap on the belts themselves but I am lately thinking of a method to have a single piece of belt for the CoreXY and clamping it at the middle somehow that it turns in two independent pieces after the tension has already been applied to the whole length. We’ll see how that goes, will involve some acrobatics as the belt is on two planes.
About your bed advise I am afraid I didn’t get it. The bed has three smooth rods that it slides up and down on and te bed plane is defined by 3 points on the T(ee) support that can be individually adjusted and can offer some cushioning in case a head crash happens. What am I missing?
About the quadrap, I was so happy when I found it that somebody else did the frame thing I was thinking about… I was looking all over the net to find the right wood construction name for the way I was arranging my “beams” as i knew that existed and planned to “name” the printer project by that name. Scout Towers came to google images search before anything else hence “Scout CoreXY”. But I am too very impressed by quadrap which has evolved since to alu extrusions and Delrin wheels … mine tries to be more basic and still has the linear rods everybody is used to… although I would have liked to have some linear guides. it would have been easier to design the XY stage.
reminds me al lot to my quadrap coreXY
I made good experience with the way you designed the bearings for the two belts on the gantry using the same axle for both bearings. To get an idea on how to implement a belt tesioning system have a look here: https://plus.google.com/111479474271942341508/posts/T6FfBRg344v
It works quite well, you can finetune the angle of the gantry to 90° this way. It’s quite simple–two bearings that can be moved in a slot, fixed with a wing nut.
My motor arangement is also like yours, this way you can minimize belt lenght, mine are about 1.8m each (print area 300x320mm)
@Daniel_F That was difficult to spot I am not familiar with the way G+ does single image albums in a post :). You are tensioning the outer belt it seems on the toothed side. I have thought of that but 1) I don’t have an easy way to do it -my outer belt quite high above the square tubing and 2) it would involve more vitamins. It is of course a very good way to tension. But I have to come with another alternative and thus far the inline tensioners are the only idea I have.
You seem to not have any printed parts, which I have galore … I haven’t managed to do away with them. But the square clamping system I have designed looks pretty straight forward to me (http://forums.reprap.org/read.php?397,587274,643926#msg-643926).
I have also designed a dual-bolt per tube-to-tube connection system that increases the surface area under which the tubes are flush with one another. I am hoping it will be more rigid that way and also doing most of the drilling with a single bit (M5). I see yours have a single bolt per tube-2-tube connection. What metric is that? Does it provide enough rigidity?
I use 1 M6 screw per tube per corner and it’s rigid enough for my frame that consits of 500mm long 20x20mm square tubes. It’s the way the 3 tubes are linked together that makes the cube rigid.
Yes that was my bet too … just thought two per “connection” would make it even stronger as the area covered by the dual bolts is slightly larger than with a single centered bolt.
It would help anyway if one of the screws gets loosened due to vibrations. It is not a costly addition so why not.
Hey guys, I took a boulder off my chest by finally taking the time/effort to re-design this #Scout-CoreXY model so that the defining components are more streamlined and repeated more throughout the model.
The new design has F697ZZ bearings instead of the non-flanche 608s so Y carriage also benefits from some weight reduction;
The new model has a slight redesign of the Z Bed Clamps to allow misalignment due to drill bit run-offs (the upper ends of the rods are straight in holes in metal);
Man it’s hard to beat this design … I can only seem to make it worse in certain aspects … but better in others :-/ . I can’t seem to be able to strike a “better in all aspects” jackpot …yet
… I haven’t managed to do away with them. But the square clamping system I have designed looks pretty straight forward to me (