Small satellites and CubeSats have established themselves within the aerospace community because of their low cost and high return on investment. Many CubeSats are developed in a short time frame and often leverage commercial off the shelf components for quick turnaround missions. With regard to the Electrical Power System, commercially available products typically use a centralized architecture. However, a centralized architecture is not reusable, since missions that require additional solar arrays or batteries would necessitate a redesign of the power system. With the range of CubeSat sizes and mission goals, it is obvious that a one-size-fits-all solution is not appropriate. This thesis details a reusable and scalable power system architecture applicable to a variety of missions. Reusability is achieved by using common building blocks or "modules," where the same modules can be used between missions. Scalability is achieved by not limiting the number of modules that can be connected together—more modules can be added as needed. In this system, solar arrays and battery units connect directly to a common bus, supplying an unregulated voltage to each subsystem. These subsystems then regulate the bus voltage to their individual needs. The power system also features direct energy transfer and solar-only operation.

The unique environment of CubeSat and small satellite missions allows certain accepted paradigms of the larger satellite world to be investigated in order to trade performance for simplicity, mass, and volume. Peak Power Tracking technologies for solar arrays are generally implemented in order to meet the End-of-Life power requirements for satellite missions given radiation degradation over time. The short lifetime of the generic satellite mission removes the need to compensate for this degradation. While Peak Power Tracking implementations can give increased power by taking advantage and compensating for the temperature cycles that solar cells experience, this comes at the expense of system complexity and, given smart system design, this increased performance is negligible and possibly detrimental. This thesis investigates different Peak Power Tracking implementations and compares them to two Fixed Point implementations as well as a Direct Energy Transfer system in terms of performance and system complexity using computer simulation. This work demonstrates that, though Peak Power Tracking systems work as designed, under most circumstances Direct Energy Transfer systems should be used in small satellite applications as it gives the same or better performance with less complexity.