With six weeks until Beakerhead, I realize the scope of the garden is far more tremendous than I can manage. I’ve decided to pull back on a couple design elements: the motors and the load platform. These were both cumbersome and complex parts that I see being the most prone to error. The load platform will also be easy to add later, if time permits.
Deciding this, I started some prototype tests. I fixed a touch sensor onto a metal pole, which will be used for activating one-off sounds. It works! As my hand came close to the pole, the sensor picks up on my presence. The pole picked up my presence at about 3cm away, which is a bit too sensitive for my liking. I’ll need to experiment with some different capacitances on C1 (the 1129 Touch Sensor comes with a 10nF capacitor), most likely lowering the value. However, the sensitivity may change once it’s installed around different objects.
For these poles, I also tried out the LEDs. I used a couple different styles of lights. First, some quite bright (3000-4000mcd) and then some less bright (500-700mcd) ones. They both work well. I put three LEDs on the end of a USB cable (for shielding from the any conductance induced by the aluminum pole), but the shared ground did not work with the 1032 Phidget LED-64. As a result, I’ll only have two LEDs on the top, which looks fine and still offers a reasonable brightness.
I’ve also laid out the board. The initial design was going to have 6 rotating boards, but after deciding to cut the motors, I can easily use one large 36″x72″ board. One benefit of the change is that it will allow more randomization of the grass blades, as there won’t be a concern of them bumping into each other when they’re swaying. Sensing movement will also be easier. With the motors, our plan was to use load cells installed on the lever extending from the motor shaft. The load cell would detect subtle compression and expansion as the blades moved the board. Without the motors, we’ll use motion sensors, which are far easier to install.
I had to go back to the drawing board to redesign this beast. Pencil in hand, I got a new chassis laid out. I maintained most of the dimensions but forwent a lot of the support that had initially been intended for the motors.
Overall, I’m liking the changes. The new design will be far easier to weatherproof. One of my big concerns with the moving boards was having rain enter between the cracks and soak the electronics underneath. While there will still be points of entry, I’m optimistic about being able to use a silicone seal to prevent excessive seepage into the internal workings.
I can’t wait to start writing the program to go with this. Once I get the prototypes done, I can work on programming something for them, which will get scaled up in the final go.
Go back and see some of the planning for the garden.