BB-8 droid remote controlled toy
I was thinking about building full-scale MO from Disney/Pixar movie Wall-E and have it ready for Dublinmaker event on 25 July 2015 in Dublin Ireland. I teamed up with my friend Mike Hibbett, an electronic wizard. Here is his website where you can get his kickstarted microcontrollers one of them so tiny, that you can embed it just anywhere for example in wearable and create the internet of things. He knows everything about electronics and programming and likes to share his knowledge in “Everyday Practical Electronics” magazine.
We started designing on a napkin the mechanics, quite complicated robot balancing on top of the ball. We even started building a prototype with 3 stepper motors and omniwheels. We found on youtube few working examples, the best was this one
We knew it will be a challenge but probably possible to make. My biggest worry was shifting the weight of the whole figure while he is moving arms and head, I wanted him to be very articulated. In the meantime, the first trailer of new Star Wars Episode 7 was released showing the new droid. It looked great and seemed to have much less moving parts. We were still at beginning of built so decided to change the project and tackle another ball robot. A quick search on youtube and we found a video of an acrylic sphere with weight shifting mechanism to control the turns:
From the video, we could see that it can provide the speed like in the trailer, but it was not clear how well it will turn. We decided to test it in a small scale so I started from scratch using cheap 10cm diameter plastic ball, one styrofoam ball, 2 servo motors, one converted to a continuous rotation and 2 micro-controllers with a cheap 2.4gh link through nRF24Lo1 boards stuck to interior 3d printed on my home made RepRap. In the meantime, the video from Star Wars celebration was released showing the real working BB8. I knew it was huge, so I assembled what I had, spent the evening painting the ball and released this poor quality video with footage of my few tests before plastic ball cracked. The mechanics worked alright, but the connection of the main axis and ball was not designed properly.
The world went crazy, everybody was searching for BB8 so my video got lots of views quickly. The announcement of the official toy made by Sphero helped a lot. Their website was showing the black silhouette with text, “this is the droid you are looking for” People started looking for it on youtube and found my crappy video 🙂
The short test proved that the ball will turn at quite wide arch. The tiny magnets will hold light styrofoam head, but it was popping very often. The biggest problem was wobble at the start and stop, but it could be probably solved with a bigger version.
The new photos started appearing online of the stage droid and static props so I prepared this set of blueprints. I noticed that details of working animatronic version of BB8 are less sharp and groves are not that deep as on the prop displayed next to Storm Troopers. It is probably to avoid edges on which dome could rapidly stop and fell off the magnetic device hidden inside the main ball. Feel free to use this plans for your own project of customized Sphero, cake, or anything you are building there. That should save you some time trying to figure out all the shapes of panels details. If you are planning something more movie accurate, look for better reference. This drawing was based on very limited reference materials at the time I was creating it and published 19 April 2015.
I started figuring out the scale of the robot based on photos from the stage where he was standing next to the R2D2. I had the approximate dimensions and even found the seller with acrylic domes online. It was costing few hundred euro with shipping to Ireland, quite an expense for a project which may never work or remain unfinished like my animatronic dragon which turned to be too time-consuming for one person with a limited budget and crappy homemade 3d printer.
I started distracting myself with other projects and abandoned BB8 idea. I still watched how others are doing, especially interested with Xrobots version 1 which reminds me my original attempt with MO project. James must be a genius, he quickly created BB8 version 1 along with other projects while producing very informative videos for his youtube channel. I am so jealous of his knowledge, skills, and talents.
I also joined the BB8 builders club on facebook where I could see the development of multiple concepts of mechanics and movie accurate droid thanks to the power of community and combined engineering skills from people around the globe. They spent hundreds of hours designing and testing and generously sharing the 3d files ready for printing in plastic.
In the meantime, I replaced my homebuilt Printrbot with budget CTC clone of Makerbot Replicator which turned to be much more reliable and more precise machine than my previous bot. I was tempted to start printing BB8 every time when I saw a Club post on my facebook timeline. The only thing which stopped me was the calculations of how much time it will take for my single machine to finish the job.
Only when I saw the modular design of the inner sphere designed by Cary Christie I realized I may speed the process. His amazing design contains 24 identical elements which joined together form the sphere. I printed this as the first part to get the sense of scale. I used minimum infill and 2 perimeters just to get the light version to hold it in hand and it looked so cool with all the ribbing and attachment points. It feels like the part of the serious machine, real robot. It took 5 hours and 30 minutes to print while proper and durable part would take around 9 hours. I would need 24 parts like that so just 216 hours of printing if everything goes well, that’s 9 days nonstop printing and I can’t leave the machine on when going to the office. I started thinking about molding and casting. It turned that I need to print 3 parts and make 3 molds to get a majority of the sphere. On top of that, I only need to fill and sand one each what is another time saver. That sounds manageable, I can give it a try, launch the printer while working on something else. I decided to print 1 ring, 1 triangle as a test, if it will be too much, I can always weather them and hang it on a wall 🙂
The whole ring is bigger than my printer’s bed so it is split into 4 separate parts and joined together with super glue. Each took 3hours to print and used 15m of PLA plastic. It was again just 2 outlines and 10%infill, much below recommended in Club instructions. This one will easily crack if put on the finished droid but I only need it as a master for my mold making, finished parts will be cast in solid resin.
They are cleverly designed with little holes, where I put short pieces of a filament and this helps to align the parts. I then sanded them with mouse sander to remove most of the layering. Next used filler putty to cover deeper lines and holes. I looked for Tamiya putty in my local model store but they didn’t stock it so I had to order it from Ebay. On the bottom, I tested universal filler for walls, but it was hard to sand and grainy. In the meantime, I bought Games Workshop liquid greenstuff which looked covered with seaweeds but did a great job. I covered the whole thing and later sanded it by hand. That started the long process of sanding and priming, sanding and priming with each pass getting closer to smooth part.
To be Continued
The story continues with triangle panels. They are made of 3 parts joined together. they fit and interlock after just little sanding. In this panel I used Bondo-like material accessible in my local shop with car accessories, it’s called ISOPON P.38 It is very easy to sand and holds very well to plastic. It smells like cancer, so I have to apply it at the very well ventilated area on the balcony and wet sand to reduce very fine dust hanging in the air. I use it to fill biggest gaps or to correct the shape. The smaller holes were later filled with liquid green stuff. Few layers of grey primer and lots of sanding gave me smooth and perfect surface.
Next moldmaking, To be Continued