University of Kentucky
Design and Implementation of Digital Logic Filtration on Open-Source Field-Programmable Gate Arrays
Grade Level at Time of Presentation
Junior
Major
Physics
Minor
Mathematics
Institution 23-24
University of Kentucky
KY House District #
79
KY Senate District #
13
Faculty Advisor/ Mentor
Christopher Crawford, PhD
Department
Dept. of Physics and Astronomy
Abstract
Field-programmable gate arrays (FPGAs) serve as extremely powerful hardware tools used in data acquisition (DAQ) and digital signal processing (DSP) environments thanks to their reprogrammable nature and parallel computation. Most often, this technology utilizes a custom algorithm of many parallel algebraic computations (some of which contain parameters that can be manipulated for desired applications) in a process called data filtration. Here I discuss the importance of DSP in physics applications and detail the development of a trapezoidal method for filtering exponentially decaying pulses on an entry-level open-source FPGA. I outline the design considerations for handling data and performing analysis, the breakdown and translation of a simple summation formula for proper implementation in digital logic, and the development and final testing of a complete project on the FPGA itself. The finalized product exhibited promising results, accomplishing nearly 98\% accuracy in most tests of pulse peak detection. Likely design flaws were also quick to be discovered, leaving an opportunity for making small improvements to said accuracy. Furthermore, there was success in generating an automated script that allows end users to recreate the complete project from the ground up with minimal experience.
Additionally, I have provided open access to the detailed development package for unfettered use and modification or application to DSP environments, with plans to incorporate new DSP projects and perform upkeep.
Design and Implementation of Digital Logic Filtration on Open-Source Field-Programmable Gate Arrays
Field-programmable gate arrays (FPGAs) serve as extremely powerful hardware tools used in data acquisition (DAQ) and digital signal processing (DSP) environments thanks to their reprogrammable nature and parallel computation. Most often, this technology utilizes a custom algorithm of many parallel algebraic computations (some of which contain parameters that can be manipulated for desired applications) in a process called data filtration. Here I discuss the importance of DSP in physics applications and detail the development of a trapezoidal method for filtering exponentially decaying pulses on an entry-level open-source FPGA. I outline the design considerations for handling data and performing analysis, the breakdown and translation of a simple summation formula for proper implementation in digital logic, and the development and final testing of a complete project on the FPGA itself. The finalized product exhibited promising results, accomplishing nearly 98\% accuracy in most tests of pulse peak detection. Likely design flaws were also quick to be discovered, leaving an opportunity for making small improvements to said accuracy. Furthermore, there was success in generating an automated script that allows end users to recreate the complete project from the ground up with minimal experience.
Additionally, I have provided open access to the detailed development package for unfettered use and modification or application to DSP environments, with plans to incorporate new DSP projects and perform upkeep.