Color-matching table and our spot at the exhibition, this is how it looked yesterday!
There were lots of people coming over curious about how it worked. And it like we put a lot of smiles on peoples faces when the cups vibrated.
It was very fun to see that most of the people trying out the cups that got matches ended up toasting with the cups after each match.. what have we really developed!? 😛
This is a rough list of resources needed to build one cup:
- 1 Thermos mug with detachable inner and outer parts (you can get one at Åhléns)
- 1 Arduino Nano
- 5 RGB LEDs
- 1 nRF radio transmitter (?)
- 4 transistors
- board with15x11 pin holes (you might want to make it slightly larger so that you get a nicer board than we did…)
- 15 680ohm resistors
- 4 1Mohm resistors
- 1 small switch
- 1 9V battery socket
- 2×15 female pinholes
- isolated wire
- Strips of copper or equivalent
- Paper cup to fit into the thermos mug.
- …and wire, solder, tools and some other bits and pieces.
The sensors have takes a lot of time for this project… We’ve experimented a lot and learned a few things along the way. The biggest issue is that we have no real ground as reference, since the cups are carried around. We’re actually using the “wrong” method for measuring capacitance, but we’ve managed to get it to work anyway 🙂
Our first problem was the the signal cables from our sensors were interfering with each other, which messed up the readings. Touching one sensor would affect all the others. We solved this by using wires with isloation, and connected the isolation to ground. In the final circuit only we managed to only have a few millimeters of wire/solder between the sensor cable and the arduino pin.
We had no way of solving the ground problem. Instead we found a work-around. By restarting the program within relatively short intervals (0.5-2 minutes) we could reset the ground reference value. I do now have the knowledge to understand exactly why this workes, but it does. We improved the readings by putting grounded strips of copper behind our copper sensors. We have some idea of why this works, but not a full understanding of it… Mostly we’ve just been trying stuff, either by reading what others have done, or just experimenting.
Lastly we moved the sensors closer to the surface to get better readings.
(Old post that was never posted)
We’ve now begun soldering our circuits onto real circuit boards. Unfortunately the first two boards were not as good as we would like – the size is OK, but the cables connected to the sensors are very sensitive to interference, so we get a lot of disturbances in the measurements. On a second iteration the sensor cables are carefully isolated up to only a few millimeters of the receiving pin, and this board seems to work even better than expected!
The design of the cup has changed dramatically from the original idea. Since we failed to find a good material for making a glowing cup we’re going for a nicely decorated cups instead. A striped pattern indicated where the vertical sensors are placed.
We reduced the number of sensors from six to four because of the interference. However, this might have been unnecessary with the success of the latest circuit. The sensors seem to be able to distinguish touch from no touch, but also an idea of how large part of the sensor that is covered! Exactly how precise these measurements are is yet to be seen.