Since my last update I have started working on another project for Random Quark alongside that which will be installed in Goldsmiths Library. This new project will be installed in a Geriatric Ward to provide a taste of nature to people confined indoors. As it stands this installation will consist of :
- A clock with a handle which can be turned to change the time in the virtual world.
- A watering can which can be turned to water virtual plants.
- Birdseed boxes which can be turned to pour birdseed out and attract birds.
These various items will likely all be accomplished with rotary encoders of one sort or another.
The first item I have finished is the clock assembly. This has been through a few iterations and I have used hardware which will allow us to keep options open in the future. I have laser engraved a light and dark side to the clock circle to distinguish each side as day and night. The rotation of the handle will be measured using a continuously rotating servo for a few reasons. First is that if we want to have the time change naturally, we can rotate the handle to keep time with the rest of the program. Using the servo also has the happy coincidental advantage that it has some resistance to it and as a result can support the weight of the handle at any angle. A standard rotary encoder on the other hand will offer as little resistance as possible, leading to a very limp handle. On the Arduino firmware side, I translated the PWM signals from the encoder within the servo into values between 0 and 360 that matched up with the angle of the handle on the clock face so that the main unity software can easily use the position of the handle.
The other two items I am still working on. The first, the watering can has proven difficult to make work in a useful way. The first prototype I built with a large watering can required the can to be supported in two places, with an axle running all the way through the can. This was problematic in the end as the screen needs to be in front of the watering can for the effect to work. I also used a very low-resolution rotary encoder so the movement of the watering can on the screen was very jerky. The next version uses a much smaller watering can and the huge weight reduction has allowed me to support it from only one side. The depth might continue to be an issue however as the linkage between the rotary encoder and the axle is still a deep 3D printed part.
Lastly the birdseed boxes, these are built and mounted to the backing piece of wood using embedded bearings. This is a technique I have witnessed on Wintergatan’s fantastic series on YouTube about building a complex music machine with thousands of moving parts. The technique worked fantastically well and has allowed me to keep the depth to a minimum so far on this piece. The boxes themselves also had a bolt embedded in them in a similar fashion in order to lock the axle in them with more than the friction of a tightly fastened bolt.
Meanwhile in the library project…
I have started building more final versions of assemblies which will be more robust and work in the final machine. My proudest is probably the tap assembly which thanks to a new 3D printed axle part is much more robust than the previous solution and uses a captive screw in the 3D print so that the handle cannot move independently of the 3D print part.
I have also started working on the labels for the final installation. We have some very interesting material that I will be using that is multi-layer. It is acrylic with a thin layer of a different acrylic on top almost like a foil. The thin layer is a different colour and can be etched away by the laser cutter leaving the colour underneath. I have been running into problems with the residue from the etching moving over the rest of the work and contaminating it with a substance which is hard to remove without damaging the thin top layer, or getting the top colour contaminating the etched parts. I expect to have to do much tape work in future to stop this residue causing such a problem.