previously: i had built my working arduino clones and was all excited to test out i2c.
so, i2c worked like a charm — with one sensor per arduino. when i was able to get a potentiometer to dim an LED up and down (when the LED was attached to the master and the potentiometer to the slave) i thought i was all set.
at home over thanksgiving, i set the system up just to make sure i had the concept down and could start programming and building my project when i came back. unfortunately, it wasn’t as easy as that. when i put a second sensor on the slave, the master didn’t know how to read it. all the readings would be overwritten by the last sensor that the master read — so, for example, if i had A0, A1, and A2 hooked up, the master would only read A2. the minute A2 was ripped out, A1′s reading would come up.
over the next couple of days, i tried everything i could to figure out how to make this work: calling the Wire.read() function multiple times, converting my readings to strings and then parsing out the data that way, etc. nothing worked, and i got more and more distressed. finally i ordered an arduino mega as a backup plan and got to work on fabricating the mannequin. at that point there was only about 2 weeks left until the presentation date, and as much as it killed me to do so, i knew i had to give up on making i2c work.
dremeling apart the doors was the most fun i’d had in a while. i got to wear a full-body hazmat-type suit and perform surgery!
hinging the doors was more problematic. i ended up using plastic toilet seat hinges and hot glue to secure the hinges and the knobs on the mannequin.
the final fabrication step was mounting the sensors, connecting the circuit board, and putting the panels into the mannequin. here are some photos:
for the processing code, i took the dummy code i had written for my ICM proposal and just modified it to reflect the sensor readings i was receiving from the arduino. i also put all my magnet switches — which were responsible for letting processing know when the doors where open and shut — on digital pins 0-7 so there was an easy way to tell the program which of the inputs were the door switches.
while i had initially wanted the whole mannequin up and running before my presentations for ICM and pcomp on 12/7, i only managed to get the doors working and one segment’s sensors working. here is a video of my progress, culled together from a 3AM video shoot the night of the 6th as well as better documentation from 2 days after i presented.
i’m still working on this project, both for the final overview documentation for ICM as well as for (hopefully!) the winter show. next steps are to implement internal code for all of the sensors in every section, as well as smooth out some of the fabrication issues i had and decorate the sensors so they actually reflect the information that’s on the screens presented. however, i was extremely pleased to hear from both tom and dan that i seem to be in a good place with this project since i do have the central control structure working. now it’s just a matter of expanding, and i’m really confident that by the time the semester officially ends, this project will be up, running, and ready to play with.