Erbium doped fiber sources and amplifiers for optical fiber sensors

This thesis explores the use of erbium-doped fiber in lasers, amplified spontaneous emission sources, and amplifiers with particular attention to applications involving fiber sensor technology. Erbium-doped fiber laser output power is shown to be strongly dependent on the erbium dopant concentration in a fiber. Using multiple fibers with various erbium ion concentrations, laser output powers are found to decrease as erbium concentration is increased. Upconversion in paired ions is successfully used to model the lasers, resulting in a better understanding of the loss mechanism involved. Further investigation shows that co-doping an erbium-doped fiber with aluminum helps eliminate upconversion in paired ions, and an optimum ratio of 20 aluminum ions for every erbium ion is established. Upconversion due to paired ions is also used to predict the behavior of erbium-doped fiber amplifiers as a function of the erbium ion concentration. With this knowledge of concentration dependence, a low doped, high output power fiber is chosen for use as an amplified spontaneous emission source in a fiber optic gyroscope. Used as a single pass broadband source in one propagation direction and as a signal amplifier in the other direction, this source is tested experimentally in a high quality fiber gyroscope. Experimental results reveal an unexpected dependence on the polarization states of the optical pump and the gyroscope output signal. A theory of polarization anisotropy in the erbium ions is developed in full and accurately models the experimental observations. Using this model to optimize the source, a fiber gyroscope output stability of 4 parts per million is obtained experimentally, approaching the requirements of inertial navigation. This model is also used to explore novel single polarization amplified spontaneous emission sources. Large scale amplified sensor arrays are examined theoretically to determine component and amplification requirements. For balanced gain and loss sections on the distribution lines, amplifier spacing that maximizes the output signal to noise ratio is found by equating the noise created by the amplifiers with shot noise. The output signal to noise ratio is shown to be weakly sensitive to the amplifier gain on the distribution lines.