The Bubble Display project involved a great deal of consideration in both the electrical and physical elements. The goal is for 64 pumps to be independently manipulated so that at any interval one can release a bubble into its respective tube. There were many variables to determine and problems to work around in the project, and discussing these were the reason for the many meetings with my advisor, Dr. Alexander, throughout the year. 

Our initial task was deciding how we were going to go about controlling 64 pumps. We decided to use an Arduino: a microcontroller that could have a user-written code uploaded to it and perform that code on whatever was attached to it. I familiarized myself with the Arduino and the coding software that comes along with it, and I was eventually able to have a single pump turn on and off at whatever interval I wanted.

We then had to decide what we could use to produce bubbles. Our two main choices were a solenoid and a standard pump. We connected both to a power supply and a tube full of water, and both yielded very similar results. The bubbles rose in a hemispherical shape, and often had a few bubbles trailing them. The only difference between the two was that the solenoid would require an air compressor and the pumps would not, so we decided to go with the simpler (and less expensive) choice of using the pumps. 

It was at this point when the medium in which the bubbles would travel through had to be determined. Initial tests with water had bubbles rising too quickly, making hemispherical shapes, and not being released as single bubbles, but groups of smaller bubbles. We then tried tests with the more viscous corn oil, and bubbles rose in a much more pleasing fashion: slowly and spherically shaped. However, corn oil has a dark yellow tint, and we wanted to have a clear liquid. It was evident that high viscosity was the variable we were looking for, so we considered glycerin and mineral oil. After a test of mineral oil, we determined it would be an ideal candidate for the liquid, as it was clear and had the performance that corn oil gave. We bypassed trying glycerin and then purchased the amount we needed from a farm supplier that had bulk amounts of high viscosity mineral oil. 

Getting 64 pumps to respond to the Arduino was the next hurdle. The Arduino only releases a 5-volt signal, and the pumps we found required 6 volts. After attempting to use an electromechanical relay to rectify this problem, we were warned that the relays would eventually break with constant use. Thus, we decided to use transistors connected to each pump to get the voltage that each needed. 

Another limitation to the Arduino is the number of output pins. There are 16 pins (places to wire up each pump) on the Arduino, and 64 pumps to control. To work around this problem, we decided to use shift registers to turn the serial output signal from the pins into a parallel signal, thus giving us the ability to control all 64 pumps with just 3 pins.   

Each element of the project is now determined, and what remains is connecting all the pieces. A stand must be built for the tubes, each 8 feet high, and next to each other are all 6 feet 4 inches long. A circuit card is also being designed to replace excessive amounts of wiring. The stand will be able to be deconstructed and rebuilt to make moving it easier, as not only will the final display be large, it will be heavy.