Contraptor

Yep! 80/20 stuff… the pallet handling and interconnect works as follows: Each "cube" is an independant unit and has a set of electrical connectors(4 large pins) and a data bus (6 smaller pins) the big pins carry up to 440 3 phase and the small ones carry EtherCAT(ethernet) so that cubes with different functions can be stacked together in any order to form a process like: pcb router —> solder screen printer —> pick and place —> heating —> cooling —> testing probe station…

Transporting between cells are pallets that the workpieces or work holders are attached to.
These are conveyed from cell to cell by linear motors in the base of each cube, on a set of tracks…
Making linear motors one of the first open source components I want to get a design together for ;-)

In the mean time I plan to build one of your kits on the CNC router/mill cell to mock up tracks, look at a placement strategy for pallet lockdowns and generally kick ideas around…

Great stuff!

James

Super cool dude! I am starting some work on the Contraptor version of a robotic arm for pick and place systems. However… I am not a robotasist so we well see how it goes. I just bought about 20 different servos to try some stuff out. I also started working on a mass produced (laser cut) version of optical encoders
to use for servo accuracy.

Perhaps we can collaborate on it.

ril3y

Great to hear you like the idea, I have been away from the project for a week

I'm thinking contraptor is the first step in a bootstrapping cubespawn set! sequence like this:

Start:
Minimum Requirements: Hand drill, miter saw, drills, (other tools) parts and raw materials…

Prefered tool set: Powered Miter saw and a drill press + other stuff

build jigs for drilling aluminum angle, drill 'em
lather, rinse repeat for other basic parts

Add cubespawn frame (12 sticks of 80/20 1010 aluminum 449 mm long(17.677"), 8 corners, 24 bolts)
Build cubespawn frame, use contraptor motion, make cubespawn parts

Finish cubespawn mill

Make a CNC contraptor set

Make Linear steppers on cubespawn

Use contraptor for custom functions within your cell-set

and someday: melt down the handmade cubespawn and the handmade contraptor sets to re-use thier materials in the furnace, re-melt, and extrusion cubes… ;-)

Now you may ask: "Why not just stick with the contraptor?" and I reply: cubespawn is what you build once you've committed to a design, contraptor is how you explore that designs specifics… different, complementary FMS tools! exciting stuff! I wish the Shan Zhai Ji (山寨机) builders would now join us and advance the design…, maybe we can contact them…

Hey James,

It's true that for the same cost, Contraptor has wider selection of components of different lengths which allows to explore a lot of different designs. I think the choice of building material really depends on what you're building. I played with several Contraptor designs in Sketchup using 80/20 instead of angle and everything fits very nicely. I used 80/20 mostly for structure and Contraptor for linear motion. One potential problem is that 80/20 is heavier than angle and depending on design/configuration, it might require more powerful steppers (which might require more powerful stepper drivers and power supply). But generally, I see a lot of advantages in having several different construction sets which can all be interfaced with each other.

On melting, one thing on my "someday maybe" list is to figure out a way to make solid aluminum objects from powder using additive fabrication + microwaving in silicon carbide, similar to this: http://home.c2i.net/metaphor/mvpage.html
The objects can then be finished with a Dremel to spec without taking a lot of time to mill.

I wonder if Shan Zhai factories use commercial pick and place equipment?

I think its important to remember that:
Cubespawn is aimed at light manufacturing,
Contraptor seems aimed at prototyping.

Cubspawn Machines are dedicated to some finished design (although it will still likely undergo refinement)
Contraptor is more flexible for the purpose of finding the "right" design (or a good enough solution to commit to)

Unless of course I'm missing the mark (always possible :-) )
you may intend to build large finished machines with contraptor components and scale up to 1 1/2" or 2" or 3"angle and all the rest…
either way I see a slightly different intent in the endpoint for the two approaches, and I think they are complementary - I really didn't consider the prototyping aspect much in the cubespawn design, just interchangability, stackability and modularity.

which is why I'm so pleased to have bumped into your project, because it DOES! ;-)

I too wonder what Shan Zhai operations look like, and what machines are needed to build thier very sophisticated products on a small scale..

I have a meeting with a vacuum forming plant owner tomarrow to pitch the cubes as a labor saving trim-out line for 14 products in his plant - I am doing it for the development progress it will bring to the project - if it works out we all get the benefit of the developement - so I may drop off for a while, but will post finished designs as they happen - sorry this is getting so far off-topic - I would like to link to your site from cubespawn and state my position there: use Contraptor to explore implementation stategies, Then populate the cube with the solution.

upward and onward!
James

I hear you, yes this is true - prototyping, bootstrapping, building makeshift robots and exploring designs are what Contraptor was intended for, so for example long term reliability was not a design goal. As you said above, the goal is to accelerate the development of a lot of things that will then accelerate the development of even more things..

When you have built something from Contraptor that you can show, please feel free to create a page in this wiki with pics/explanations of your project, we'll definitely link it in the Contraptions page and let people know about it. Also, we can link from home page to your site if you're looking to attract contributors to your project.

RE Shan Zhai operations - Google is our friend:) http://www.shanzai.com/index.php/market-mayhem/asia-analysis/270-shanzhai-factory-photo-tour

That sounds like exactly the goal: we can do a little to help promote each other's sites, as we promote our own, thereby accelerating the whole idea a little faster, and I'll do as you say: post designs that I'm using Contraptor to define or support. Hopefully we'll get people from both camps using the whole toybox to concieve, design, and then build and automate thier creations.

Oh cool. I was thinking of removing the IR filter from a green pointer laser and trying that, but a laser already made for burning is even better. I actually haven't taken the dive into researching the availability of burning lasers, since I figured I'm make the cnc first.

I've thought about trying to combine the beam with a prism too. Then you could just get a bunch or IR laser diodes and set up something that could cut balsa.

Hotwire cutter should be relatively easy to do. What are the typical speeds of moving wire through the foam?

I haven't looked up the speeds, but i'm assuming you could easily use a belt drive, but the precision of a leadscrew could be important depending on what you're casting.

Actually, now that I think about it, I think a router bit would be a better option for foam than a hotwire (unless you really cared about the speed). High density insulating foam seems like it would be a perfect material to mill through as long as you kept the router bit slow to avoid glazing.

There are couple of scenarios for cutting stuff out of foam

- model car shape - router is better
- wing from a long/huge piece - hotwire cutter is better

With linear bearings, belt drive should be pretty accurate, but I think even 1/16 should be good enough in most cases of cutting something large from the foam?