Solar-Powered
Phone Charger

Frances and I investigated how to power a phone from solar panels whose output changes with sunlight, temperature, and positioning. We compared regulator designs in PSIM, calculated component requirements, built a soldered prototype, and tested the complete charging system.

I worked on circuit design, soldering, PSIM simulation, and design iterations, and presented our findings to Northeastern faculty and mentors. Our research was guided by Karen Abbaskhanian and Dr. Mahshid Amirabadi in the Power Electronics and Renewable Energy Systems Laboratory.

We compared three views of the solar panel’s current–voltage behavior: the manufacturer’s curve, a curve generated from PSIM data, and one built from our experimental measurements. These gave us a way to examine the source before choosing a regulator.

We studied a switching buck converter, a resistor divider with a Zener diode, and a series-transistor linear regulator. We used PSIM to compare their behavior and calculate the component requirements for a 5 V output.

Buck converter

We calculated a duty cycle of 0.27 for an 18.5 V input and a 5 V output, then combined the solar-panel model with the converter in PSIM.

Zener divider and series transistor

The Zener-divider approach required more current than our selected diode could support, so we did not use it for the prototype. For the series-transistor design, we modeled varying input-voltage cases and calculated the current and power required by the reference network.

We selected a series-transistor regulator with feedback for our physical prototype. Its feedback network and error amplifier adjust the transistor’s operation to regulate the output as the input or load changes.

We built and soldered the circuit, then tested it with a DC source, a variable load, and output-voltage measurements.

A stable voltage did not immediately make the phone charge. We investigated the USB-to-Lightning interface and added resistors to provide the signals the phone needed to accept charging.

In outdoor testing, we connected three solar panels in parallel to supply the current needed by a phone. Our final output-voltage measurements showed less than 1% error.

We also found clear limitations: the transistor overheated, and the regulator’s dissipative behavior meant we needed three panels. We did not complete the other planned linear-regulator builds because our components did not arrive.

Our technical slides document the regulator comparisons, calculations, simulations, interface troubleshooting, and testing results.