A continuation of Emerick Gilliams' project, linked here. This project seeks to extend and complement the work done by Emerick. The primary extension in this project is to provide the sensor system with a self-sustaining power source using the fuel cells. This system should allow for nearly indefinite periods of time, likely to be limited by health of the battery rather than simply the charge. In addition, more information will be gathered on the performance and characteristics of the BMFC including polarization curves and power density curves. The primary focus for the electronic and code design will be power efficiency and modularity.

While still in development, the heart of the system will (likely) be a Ti TPS61200 based DC-DC converter. This device was choosen due to its low start-up voltage (and lower operational voltage) which complements the capabilities of the fuel cell well. Another option would be based on the LTC3108 however, with a max V_in of 500 mV, the fuel cells we tested would have been incompatabile without more electrical components. The rest of the the electronics consist of a microcontroller and a LiPo Charger

The microcontoller powering the sensors is currently an Arduino Pro Micro, although this may be swapped for a more power efficient processsor down the line. The board is interfacing with a RTC and an SD card, in order to properly timestamp and log measurements. Voltage measurements are taken using the Arduino's built in Analog to Digital Converter(ADC). The ADC provides 10 bit resolution which gives a voltmeter resolution of 2¹⁰ steps. Since the reference voltage is 3.3 V, the voltmeter can measure in approximately 3 mV steps. The microcontroller should be set to sleep most of the time, waking only to measure and record data. Utilizing this functionality, power draw can be significantly decreased.

The star of the show. The fuel cells in use are Benthic Microbrial Fuel Cells(BMFCs). These fuel cells take advantage of microbes living in marine and freshwater sediment to generate electricity. Unfortunately, the voltage produced by these cells is in the hundreds of milivolts. This makes direct connection to most sensors impossible, and due to the nonlinear nature of the fuel cells, typical series and parrallel structures have not been a typically adopted solution. They also have relatively modest power densities on the order of 10s of mW*m^-2, meaning that any sensor needs to be incredibly low powered in order to not outpace the fuel cells power output. On the flipside, fuel cells have proven to be reliable longterm power sources for sensor networks.

While most BMFCs have been deployed in marine or wastewater enviroments, the fuel cells in question will likely be deployed in freshwater like Lake Okeechobee: This comes with its own set of benefits and drawbacks. For one, the lake is freshwater, which is less conductive, but also less corrosive. A bigger deal may be how affected everything is by the wind, which will increase the homogeneity of the lake, as well as the chance of a fuel cell becoming dislodged. Finally, Lake Okeechobee also has some concerning ecological concerns meaning that longterm sensor data could provide answers on how to best deal with these problems.