Prusa Vertex vs New (hacklab) Vertex
A couple months ago, myself and some other hacklabers branched the Prusa Mendel RepRap and started working on it. I wanted to post about a cool feature Rob and me have been working on.
Well, its actually 3 interconnected features. The first is that configurations for the hacklab reprap include configurability of the build volume:
// Build Volume Settings //================= //The x,y and z printable dimensions in that order build_volume = [100,100,80];
Of course, vertices had to be redesigned to accommodate different x:z ratios. When making the new parametric designs, we redesigned the entire thing. In the process, we reduced size, reprap size and (naturally) build time.
Comparison of the new (right) and old (left) verteces at a joint.
Notice not only the size difference but how the old one takes up space by leaning back. This is especially noticeable if, like Rob, you are making a miniature version.
Rob's Mini RepRap with old parts.
And of course, the old one can’t be an arbitrary angle.
And I think this configuration emphasises the size difference:
Two pairs of the new vertex and the old one
And that’s before we compare against the Mendel vertices.
Look at how much thicker that thing is!
In fact, the new ones are so much smaller than the old ones that four can fit on a single ABP (BTW, thanks to makerbot for giving Rob the freebie! They’re pretty awesome, and having a loose one made it a lot easier to figure out how to fit them on the platform.)
4 new vertices on an ABP
I’m pretty sure that you could fit another vertex on there. As for the top… Well lets look at the original set up
Old top vertex with z-motor-mount.
Originally, the top was three separate pieces. This really didn’t make sense, so I made it into one piece.
The new top vertex.
The new top vertex is actually just our new version of the z-motor-mount part. Since it integrates 3 parts in to one and there are two instances of it in the reprap, it cuts the part count down by 4 parts. The above images is with a vertex angle of 80 (and thus the top is 20), so it also exemplifies the new configurable angle set up. Yay openscad!
Here is a 60 degree one on my RepRap:
60 degree z-motor-mount on my RepRap
We have exciting plans for the future. Just in this stream, we’re planning to have the build process include custom documentation for your configurations (including the rod lengths necessary to get the build volume you want) and 3d models of the assembled RepRap with your specifications. Look at our github repository to see the other cool features we’re working, including single top rods, threaded rod instead of belts and lots of other cool things!
Tags: 3d printing, openscad, reprap, vertices
January 9, 2011 at 11:25 |
Wow, nice work!!!
January 9, 2011 at 14:00 |
Thanks.
January 9, 2011 at 13:39 |
Nice work. Are there any plans to make the hole sizes parametric so that we backward Americans can make a SAE version. SAE vs. Metric here is a hugh difference in price for the vitamins!
January 9, 2011 at 14:27 |
Not only are there plans, but it is already done for many parts. For example, not only are vertices parameterised by threaded rod thickness, but also nut thickness and diameter.
From configuration.scad:
// Threaded Rod Settings //================= //the diameter of the threaded rods threaded_rod_diameter = 8.4; //the radius of plastic needed around each threaded rod hole //to hold it in place threaded_rod_horizontal_clearance = 4; //the plastic needed above and under each threaded rod hole threaded_rod_vertical_clearance = 2; nut_diameter=12; nut_thickness = 7.8; tear_shape = false; //Use tear shaped holes for rods? not compatable with tight_grip = true tight_grip = false; // Use tight holes where possible so that parts will self tap when // threaded rod is drilled in. Reduces need for nuts. // tight_grip = false not compatable with two_top_rods = falseJanuary 9, 2011 at 17:27 |
Also with the current settings in configuration.scad the design is metric/imperial agnostic because m8 is almost identical to 5/16 threaded/smooth rod. And with the current vitamins that is the only part that would change for imperial/metric.
Actually, a point I think we forgot to mention was that the hacklab reprap vitams list is quite small and easy to source atm:
– threaded + smooth rods
– zip ties
– extruder + electronics
– teflon sliders [optional, more on this later – as in once I’ve finished designing the parts!]
January 9, 2011 at 18:04 |
is the stiffness of the machine impacted at all?
January 9, 2011 at 19:53 |
The new vertices seem much stiffer. I’ve found my frame to be much more sturdy since i switched to the new vertices. But that’s just my subjective impression. I don’t have any electronics on it yet and haven’t done any rigorous tests.
January 10, 2011 at 11:25 |
I never understood why the upper vertices weren’t done that way all along. As for the lower vertices, I like the idea, but I think there’s a simpler way to do it without that inserted nut. It’s a little hard to explain without pictures, but I’ll try to get it mocked-up later today.
January 10, 2011 at 19:10 |
Here it is: http://www.thingiverse.com/thing:5515
January 10, 2011 at 20:46 |
Wow! That’s a cool design!
It has several down sides:
– It has a greater volume, increasing build time.
– It has a greater x length, increasing foot print.
On the up side, it is a simpler design.
We’re really optimising for foot print and build time, but if you want to redesign this to be compatible with our horizontal bars (see common.scad: frame_horizontal_struts ) and support arbitrary angles, I’d be happy to pull it into our repo as a possible configuration. Are you on github?
EDIT: Also, having the nut inside isn’t completely a downside. While it increases complexity, it seems to help it reduce vibration.
January 10, 2011 at 21:20 |
Arbitrary angles are already in there. The horizontal bars will be harder, because the crossing bars have to be able to pass between them (it will limit the range of the angles).
Not sure about the volume part, but its a simple enough shape that it could be shelled without much impact. The X (and Z) lengths are greater, but because of the reduction in the Y, you could fit 6 of these on a makerbot platform with enough room left for a row of bar clamps, whereas yours can only fit four. Or, since you only need four, you could probably fit two of them, an upper vertex, and at least four bar clamps (especially if you use my original V2 clamps, which are a bit smaller and align a little better, instead of Prusa’s). That’s basically the whole frame in two Makerbot plates.
That vibration reduction claim sounds fishy. The only reason I can think of for that to happen is because there is very little material between the nuts that connect the angled rod, which means it would have more play.
January 10, 2011 at 21:21 |
Oh, and yes, I have a github account, but I haven’t really figured out how to use it yet.
January 10, 2011 at 20:54 |
Thanks for reduced part count and the cool design ideas. I’m close to start printing a Hacklab.to Reprap myself.
But I’m not fully convinced the old vertex used up space. Essetially it saved space (at the very corner, which isn’t included) and rod. Because even if you had a zero volume vertex – you couldn’t use the full rod length, because whatever is moving along the rod – it couldn’t go all the way to the end….but maybe I’m not entirely right here. Still confused….
I like whosawhatsis’s idea – not sure it would work for all angles….
January 13, 2011 at 00:45 |
Actually that’s all dependent on the y-carriage. As long as you raise your build platform a bit above the bars and their frame attachments then the build platform can travel past the end of the Mendel frame. The limiting factor in that case is the size of the y-carriage the build platform is attached to which can be made considerably smaller.
January 13, 2011 at 06:44 |
Your Y travel is still limited to the length of the Y bars minus the length of your froglet, though. If you want to increase your platform size without changing your frame size, you have to make your froglet that much smaller, and you loose stability if your platform gets too much bigger than your froglet.
January 24, 2011 at 18:27 |
[…] Christopher Olah's Blog « Hacklab RepRap Coolness […]
March 6, 2011 at 20:48 |
Question: are you still maintaining the design on git? I was looking to build your vertexes, but the latest commit appears to have some critical problems.
I went back to commit a1e2ab, and it builds properly…
March 6, 2011 at 21:58 |
I haven’t done anything recently, but the hacklab-reprap project was updated as recently as 5 days ago. Are you looking at my repo or the hacklab one (https://github.com/hacklabto/Hacklab-RepRap)? And are you on the main branch?
I’ve been busy, but should be doing a lot of work on reprap in the near future.
March 7, 2011 at 18:34 |
I’ve been looking at the hacklab variant (the one you linked). The latest commit I see, on the master branch, >40 days ago (1/26/11).
I decided to pull design intent from the a1e2ab commit and then build from the ground up in SolidWorks (where I am most comfortable/proficient). I was able to further reduce volume in the feet by 11.6% while adding some general ‘good engineering’ additions. You guys did a great job at optimizing for volume – I’ve got a couple more tricks to incorporate later.
I also remade the upper vertexes to less than half the volume of the original Prusa design – I’m considering the merits of combining into the motor mounts as hacklab did. There’s def. a benefit to having them separated (particularly if machine size is ill defined).
Next, I’m working on the other large volume parts.
March 22, 2011 at 07:22 |
It seems likely that your problem is that mcad didn’t get pulled. Check if there is anything in lib/mcad and if not, pull the hacklab mcad version into it. Sorry about taking so long to respond…
Also, I’d be very interested in hearing more about your work. Write a blog post? 🙂