Honors Theses

Date of Award

8-2025

Document Type

Undergraduate Thesis

Degree Name

BS

Department

Electrical and Computer Engineering

Faculty Mentor

Samuel Russ

Advisor(s)

Edmund Spencer, Mohamed Shaban

Abstract

This project presents the development of a high-performance, cost-effective, and powerefficient pipelined architecture designed to execute 10,000 Fast Fourier Transforms (FFTs) per second on an Efinix Trion T120 FPGA. Each FFT processes 4096 signed 13-bit elements, facilitating real-time data analysis for a Time-Domain Impedance Probe (TDIP) operating at a maximum data rate of 490 Mbit/s. The algorithm utilizes the built-in multipliers and dual-port memory cells of the FPGA to optimize data storage, transfer, and processing. It achieves this high performance while only using approximately 4% of the Look-Up Tables (LUTs), 7% of the integrated RAM cells, and 15% of the multipliers available in the FPGA. The Verilog implementation successfully performs one FFT in 6216 clock cycles with a maximum clock rate of 58.8 MHz, permitting 9459 FFTs/second/ module.

Share

COinS