Crosstalk and coupling in reflectometry testing of electrical wire faults
Lucas R. Thomson
T h e U n i v e r s i t y o f U t a h G r a d u a t e S c h o o l STATEMENT OF THESIS APPROVAL The thesis of Lucas R. Thomson has been approved by the following supervisory committee members: Cynthia Furse , Chair ABSTRACT The focus of this thesis is the impact and use of crosstalk and coupling when testing for electrical wiring faults using reflectometry. This thesis describes a method for detecting and locating faults on cable shields using an adapted reflectometry system. A signal
... on the inner conductor is coupled to the outside through the fault, a small aperture in the cable shielding. This very small signal is then detected and correlated with the original signal transmitted on the inner conductor. The signals that leak out of the aperture, the damaged shield, and propagate down the outside of the cable are quantified as a function of the aperture size and frequency. A ferrite loaded toroidal sensor design is also proposed for receiving this external signal in order to both detect and localize the shield damage. Both simulations and measurements validate the effectiveness of this method. Unshielded discrete wires are another common type of transmission line. While unshielded wires are primarily used for DC power, they are still subject to degradation over time and require maintenance. Unlike shielded cables, there is a significant amount of coupling that occurs between adjacent wires during a reflectometry test. This coupling is quantified and evaluated for two applications. The first is simultaneous testing of multiple adjacent wires in a bundle. In this case, minimizing the coupling is desirable in order to reduce noise in the reflectometry signature. The second is the exploration of the potential for a single reflectometry test to locate faults on adjacent wires without directly testing them. When a single test is performed in a multiwire bundle, the reflectometry signature will be a superposition of reflections from all nearby conductors. This thesis addresses the testing of a multiconductor wiring structure with a common signal reference as well as a similar structure with an isolated signal reference. In order to accurately detect faults on multiconductor wiring structures, both testing methods must be considered. A fault between a conductor and its reference conductor is easily detectable. A cross fault between two nonreference conductors is not. For cross fault consideration, the only method for detection is using a common signal reference and analyzing the data on adjacent lines.