Sigma-Delta/FPGA Accelerated Controller Design

Digital controllers, or devices that regulate system behavior in many electrical and mechanical systems, have become widespread and varied in function in the modern era. Though controllers exist in many places and for many applications, many controller designs suffer from multiple problems, primarily high latency, low resolution, high power, and high cost, which creates a cost, efficiency, and/or performance barrier for many applications. It is the job of engineers and researchers to find and create controller designs that seek to reduce the severity of these primary issues and be able to create designs that meet the demand of a system. In recent history, researchers have found a possible solution by developing controllers adapting currently available technologies: sigma delta converters (SDCs) and FPGAs (programmable circuit logic). SDCs, though primarily developed for converting analog signals like audio into digital data, along with FPGAs hold potential due to their ability to output computations at speeds much faster than many hardware and software solutions while maintaining low power, high precision, and low cost. Though they hold much potential, substantial work needs to be done in terms of designing controllers with these technologies and observing their computational behavior and stability.

Current research under the guidance of Professor Forrest Brewer at UCSB aims to design a foundation to test SDC/FPGA designs and examine their computational behavior. The current project studies an FPGA/SDC based multi-phase oscillator which has practical applications such as driving motors and performing audio techniques like audio output and cancellation. The framework developed for studying this design will prove applicable for similarly based controllers in regards to how computational and power behavior is observed, how to structure designs and obtain data for other applications, and how to improve future designs. Findings and the framework created within the last 6 months have already proven useful for other projects and have been distributed to both graduate and undergraduate (capstone project) students. The future goals of this project are to complete the current framework and finalize the design to do an intensive study of computational behavior and apply the oscillator in useful applications.