We were super sick of building a new structure, base, support, etc, every single time we wanted to test something! And building something and then realizing what needed to be changed. So, we decided to stop doing that. We wanted to make a machine housing that was easily adaptable. We wanted to be able to quickly pull it apart and put it back together (so definitely no glue – glue is the enemy). We also wanted to easily be able to add attachments without having to go find a drill, and measure, and find material to add on, and make it stable, etc etc etc. Seriously, we were so sick of this.
We also really wanted to be able to do many tests in a row, cataloging results in a methodical manner and having easy access to change variables.
Soooo…. we made a MakerBot!!!
Just kidding. Seriously though, the next person who asks us about the 3D printer we are building will be sorry. Having a very public space in studio is just us repeatedly answering questions for prospective parents on their tours through. And pointing out that there are four 3D printers in the studio, but this is definitely not one of them. It’s really just a box. A super adaptable box, but a box.
In all transparency, the “build your own 3D printer revolution” was in the back of our minds while designing this. And obviously with good reason, because it works!
So we built a box, from laser-cut double-coated masonite from Lowe’s and variations of nuts and bolts and brackets, but all #6-32. There was no way we were going to make our lives more difficult by using more than one hardware size. There is also a piece of 1/8 inch acrylic that we had laying around, and that seemed like a good idea. We wanted to make “trays” or various beds that could be easily interchanged, and being able to see the variations seemed like a useful idea.
We went with a 15 inch length in all three dimensions, partially because it would give us breathing room for a 12inch bed space, and partially because we had 32×48 inch pieces of masonite, and those could easily be cut into thirds at 16 inches, and fit in the 18×32 inch laser cutter.
Problems with V1.1
We didn’t have a good way to attach the pendulum to the top besides tying the string through.
Problems with V1.2
Spheres spinning, pendulum attachment, not enough magnets, didn’t get the spring joint together
Switched to 1/4″ x 1/4″ cylindrical magnets in the bed. We attached a 3/4″x1/4″ cylindrical magnet with a #6 screw end on the end of our pendulum and added a second magnet of the same size to the end of it to increase the pull force of the magnets. We originally used a 1/2″ dowel, but switched to a 7/16″ when the tolerance on our 3D prints wasn’t perfect.
The pendulum is attached to a gimball like joint that will allow it to rotate freely within a single axis. Weights are attached halfway through the dowel (above the joint) which allows for constant motion of the pendulum.
There is a variation of bottom tray template that we created. By having different geometries and positions of the magnets we are hoping to achieve many different variations of motion that stay within a constraint that we have control over.
Below are the 3D printed pieces that serve as (from right to left) case for the light source, weight attachments and gimball joints.
Problems with V2.1
The dowel was so annoying to work with. Constantly drilling and un-drilling our screws to make minor adjustments did not make any sense. The dowel was getting weaker each time and it was impossible to be precise. The dowel was also a light color, which we thought would affect our photos. Even though we planned to paint it, we eventually decided to just do away with the dowel dynamic. We created a systemic approach to the construction of our machine to avoid doing stuff like this, and we remembered why it mattered so much.
The screwing of the 3D prints to the dowel was also affecting the weight balance of the pendulum, and although we tried to alternate directionality to accommodate for this, it wasn’t a perfect fix. Screwing in and out was also slightly degrading to the 3D prints, as it was very difficult to do it exactly accurately every time.
The gyroscopic joint wasn’t great. It worked, but the spacing between the brackets was too large and made it sag, which made the whole system less precise. It also looked seriously sad. We were also SUPER annoyed that we designed it to connect directly to the box, so that we couldn’t just pull the pendulum off whenever we wanted. We love this about Version 1.2, and we wanted to go back to it. It was seriously so convenient.
We also realized that the weight at the top was not the correct move, and that with the repelling magnet forces, we wanted the weight at the bottom to keep the pendulum moving. We also noticed that a very little bit of weight made a difference, so we decided not to use the 1-1/4 sport weights, and make rings from masonite that could be stacked at the bottom of the pendulum. We also wanted to be able to control the amount of weight with a smaller degree of change.