Skip to main content
SHARE
Publication

Signal Processing of Multiplexed Optical PWM Signals for Sensor Arrays in Nuclear Environments

by Daniel C Sweeney, Frederick K Reed, Kathleen C Goetz, Nora D Ezell
Publication Type
Conference Paper
Book Title
13th Nuclear Plant Instrumentation, Control & Human-Machine Interface Technologies (NPIC&HMIT 2023)
Publication Date
Page Numbers
1642 to 1650
Publisher Location
Illinois, United States of America
Conference Name
13th Nuclear Plant Instrumentation, Control and Human-Machine Interface Technologies (NPIC&HMIT 2023)
Conference Location
Knoxville, Tennessee, United States of America
Conference Sponsor
American Nuclear Society
Conference Date
-

Safe and effective generation of terrestrial nuclear power greatly benefits from the actionable data provided by the array of sensors located throughout a plant to provide a holistic online indication of reactor operation. This array includes sensors for monitoring coolant flow and pressure, temperature and heat transfer, radiation levels, structure health monitoring, and other critical parameters for reactor operation. While sensors capable of measuring these parameters have been developed, the electronics used in the pre-amplification and analog-to-digital conversion of the small signals they produce are extremely sensitive and susceptible to damage by high-temperatures and radiation environments nuclear reactors encounter while generating power. The small signals from sensors in nuclear power plants (NPPs) are transmitted over long cable runs which introduce dispersion artifacts into the signals of interest as well as electromagnetic interference (EMI) from lighting fixtures, pumps, mains electricity, and other equipment. To overcome these challenges, a front-end digitization (FREND) platform has been developed to use radiation-tolerant electronics to multiplex, amplify, and optically encode signals from an array of sensors for transmission over an optical fiber to mitigate dispersion and EMI artifacts from long runs of electrical cabling. To recover the optically transmitted data, a signal processing scheme based on 1-dimensional template matching to an indexing channel is described herein and demonstrated to have an effective bit-depth of 9.2 bits (1%). This scheme has been validated in proof-of-concept, non-nuclear testing and preliminary experimental results show good agreement between the measured optical output and and sensor input signals. The FREND system represents a low-loss data link between sensors in nuclear environments and data acquisition hardware which is aimed at improving the signal-to-noise ratio of the data acquired from these sensors to provide better information to operators and researchers.