The need to miniaturize traditional optical devices and to incorporate them into a fiber environment has led to the study of in-fiber lasers with doped fiber cores and distributed- feedback grating cavities. This paper implements a recursive Green's function approach to model the threshold gain for distributed feedback in-fiber lasers. It is shown that the Dyson equation based model, which was originally applied to diode lasers, can also be used to model distributed feedback fiber lasers. A one-dimensional recursive Green's function process using Dyson's equation is used to obtain threshold gains and resonant frequencies for in-fiber, distributed feedback lasers. The use of Dyson's equation allows exact solutions to arbitrary accuracy for any grating shape. The recursive Green's function process allows rapid calculation of periodic or uniform structures of any size, yet does not easily yield the electric field values inside the grating structure. Also, the recursive method is easily applied to chirped gratings and phase sections between grating pairs, and can be extended to higher orders.
We report on the realization and investigation of a novel `continuous fiber' device structure whose spectral response is that of a comb-type bandpass filter. Filter response was investigated as a function of overlay refractive index and thickness for both oil and solid-state overlays. Passband linewidths (FWHM) of < 5 nm, insertion losses of 0.5 dB and free spectral range 25 - 250 nm have been demonstrated. The filter response was tuned over a 45 nm range by variation of the overlay superstrate index. In addition, the use of lithium niobate as the overlay offers the potential for electro-optic tuning and this topic is currently being investigated.
A novel `continuous fiber' tunable bandpass filter for application to the wavelength selection and tuning of fiber lasers is described. The characteristics of an erbium doped fiber ring laser incorporating such a filter are reported.
We report a new theoretical study of the effects of anisotropic erbium dipoles on the polarization of Er-doped fiber lasers (EDFLs), as well as measurements of the polarization of EDFLs made of a polarization-maintaining fiber. Convenient closed-form expressions are presented for the gain of a signal polarized either parallel or perpendicular to the pump. These results are used to interpret the polarization behavior of our EDFLs, stressing the effects of pump orientation, fiber length, and cavity Q. We identify in particular operating conditions that produce a fiber laser with either a linearly polarized output or equal power in both polarizations.
A theoretical method is presented for investigation of modelocking in a fiber laser, where the modulation is performed by the cross phase modulation (XPM) from synchronous pumping using an external signal. Using this method, modelocking in a cross phase modulated fiber laser has been investigated.
Fiber optic channel equalizers are devices of prime importance in multi-channel telecommunication links and networks. They are used to compensate for the channel power imbalances accumulated along amplified long links and stabilize the channel optical powers. In multi-stage amplified optical links, channel power imbalances occur as a result of two factors. Firstly, the emission and absorption cross-section variations across the erbium-doped fiber amplifier (EDFA) bandwidth result in channels experiencing different gains and acquiring unequal output power. Secondly, the output power imbalances are further deteriorated by the fact that the Er3+ transition is predominantly homogeneously broadened at room temperature. Therefore, the strongest signal (channel) saturates the gain medium and compresses the gain uniformly at the expense of the power of the weaker signals. To avoid the build-up of channel power imbalances and the subsequent detrimental effects on signal-to-noise ratio in multichannel optical links, various channel equalization schemes have been proposed. They include the use of Mach-Zehnder optical filters or acousto-optic tunable filters to selectively attenuate the higher-gain channels using active servo-loops. Dynamic channel equalization can also be achieved by controlling the gain, by adjusting the pump, in a two- stage amplifier with complimentary gain spectra in each stage. Also, passive channel equalization has been demonstrated by cooling the amplifiers down to approximately 77 degree(s)K so that they become predominantly inhomogeneously broadened.
An erbium-doped polarization-maintaining (PM) fiber has been manufactured. The core is codoped with Ge and Al and a gain of 36 dB is achieved. The fiber design utilizes the combined effects of an elliptical stress-inducing cladding and an elliptical core to achieve a high birefringence of 3.1 X 10-4 and a polarization crosstalk below -37 dB for a 10 m length. To optimize the birefringence, a model of the distribution of thermal stress in an elliptical-cladding fiber has been developed, which yields an accurate expression for the birefringence. The additional geometrical birefringence due to the elliptical shape of the core can be superimposed on this result. Measurements of birefringence versus wavelength confirm the predictions of the model. The erbium-doped polarization-maintaining fiber has been tested in an all-PM fiber ring laser having highly linearly polarized output.
This paper is divided in two parts. The first one is dedicated to the noise figure computation of noise saturated EDFA. In-line and preamplier EDFA are mainly concerned. The noise figure calculation requires the computation of the whole ASE spectra. A discretization of the fluorescence spectrum is detailed in order to rapidly and accurately compute the EDFA noise. The gain and noise values obtained with our method have been compared to those of various European research laboratories: there is no meaningful (< 1%) differences between our results and methods which consider the fluorescence at every nanometer. The same numerical accuracies are obtained leading to smaller computation times. In the second part, the noise figure of booster EDFA is investigated. In the case of booster EDFA, the input signal is quite high (a few mW). The laser levels populations are dependent on the photon number. The now classical noise figure formula have to be carefully applied because of this dependence. A direct integration of the master equation is required to compute the real noise figure of power boosters. The discrepancies between both definitions have been quantified: differences up to 3 dB have been computed for a 980 nm-forward pumped EDFA.
In many amplification regimes, the EDFA gain is saturated by noise. This is particularly true for in-line and preamplifier EDFA. The accurate computation of noise spectra is necessary to predict the EDFA exact gain and noise characteristics. In most cases, this is done by solving the rates equations describing the noise spectra. An interesting approach of noise unsaturated EDFA has been given by E. Desurvire. One ordinary differential equation (ODE) describes the whole noise spectrum in one direction. This is the usual concept of noise effective bandwidth (NEB). A set of 4 ODE is then required to compute gain and noise characteristics, leading to short computation times as a small size ODE system is concerned. The NEB definition we used has been modified in order to take into account the spatial distribution of the light power in the active fiber section and the population inversion. The NEB method we propose uses 1, 2, or 4 NEB, leading to small computation times (approximately equals 4s.). In order to generalize the results obtained by our method, several cases have been investigated: input signal powers of -40, -20, O dBm and input pump powers ranging from 10 to 30 mW. The worst case is the low input signal power because of the high noise level saturating the EDFA gain. The predictions provided by one modified NEB are at least better than the results obtained with 4 standard NEB, as defined by E. Desurvire. The computations fairly agree with the reference solution (integration of the rate equations): the larger discrepancy is lower than 0.7 dB.
Recently, there has been much progress towards the establishment of SiO2 as a practical non-linear material. Three processes currently being explored for creating second-order nonlinearities in SiO2-based materials include: photoinduced second harmonic generation observed in fibers and fiber preforms, temperature/electric-field poling of bulk fused silica and of SiO2-based waveguide structures, and electron-beam irradiation of bulk lead silicate glasses. This paper reviews the current state of these three processes and discusses some advantages and disadvantages.
We report an all-optical switch in a two-mode silica fiber based on a resonant nonlinearity introduced by color centers in an irradiated P2O5-SiO2 fiber. Full switching of a 900 nm signal was observed in a 22 cm length of fiber for only 1 (mu) J of absorbed 532 nm pump energy, with a projected response time under 10 ns. This switch operated with a peak power two orders of magnitude lower than for a comparable Kerr effect device operated under identical conditions. Photobleaching was observed and is believed to be a solvable issue.
Phosphosilicate and germanosilicate optical fibers doped with Tb3+ are darkened rapidly by exposure to 488 nm light, resonant with the 7F6 yields 5D4 transition of Tb3+. The induced absorption decreases monotonically throughout the visible, becoming negligible at around 750 nm, and can be bleached out by exposure to 514 nm light. In phosphosilicate fibers the bleaching goes virtually to completion, but in germanosilicate fibers it is hindered by the gradual formation of Ge-based color centers. The darkening and bleaching processes occur over comparable timescales, and require three and two photons respectively. For the optical intensities used, the bleaching of the induced loss is photonic, not thermal in nature, but thermal processes may become important at higher intensities. Other argon laser lines are able to induce darkening to a certain extent, dependent on the overlap with the 7F6 yields 5D4 band, and the equilibrium absorption state of the fiber depends on the competition between the darkening and bleaching processes. The absorption state, once established, is stable at room temperature over a timescale of months. The darkening is believed to occur by photo-ionization of Tb3+, with trapping of the released electron at sites in the network; the absorption may be due either to these electron traps or to charge transfer bands between Tb3+ and Tb4+. Refractive index changes associated with the absorption changes may be useful for side-writing fiber gratings with blue-green light, and the reversibility of the absorption changes may have application in optical memories.
We present a theoretical model that computes the nonlinear index (n2) of semiconductor- doped glasses (SDG), based on the material's properties, and predicts the power and length requirements, as well as the optimum operating wavelengths, for an all-optical SDG waveguide switch. The main conclusions are that (1) n2 depends strongly on pump intensity, which partly explains the large disparity in reported values of n2, (2) the pump and signal wavelengths should be in specific and different ranges to minimize switching power and signal loss, (3) for CdSSe- and CdTe-doped glasses, n2 is relatively small, and the switching power requirement for these two SDGs is consequently quite high (2 - 16 W). We provide evidence that this weak nonlinearity, compared to that of similar semiconductors in bulk, is due to the strong nonradiative recombination of carriers arising from the small size of the semiconductor microcrystallites. Projections indicate that the switching power would be reduced by up to three orders of magnitude by increasing the microcrystallite size, thus producing a slower (ns) but more power-efficient switch.
A fully connectorized diode laser pumped first window amplifier has been constructed, for the first time, around a thulium doped fluoride fiber. For a 780/806 nm combination of pump and signal wavelengths intrinsic small signal gains of 25 - 26 dB and gain efficiencies of 2.4 dB/mW have been achieved. In addition output powers approaching +13 dBm from the doped fiber were possible for launched pump powers of 31 mW, which corresponds to a pump to signal conversion efficiency of around 64%. Minimum noise figures of 4 - 5 dB have been achieved with characteristics which mirror the output power performance. The 3 dB gain bandwidth for this transition has been measured at 9 - 10 nm for small signals, increasing to better than 15 nm under signal saturation conditions.
Amplification characteristics of graded-index (GI) type organic dye doped polymer optical fiber amplifiers (POFAs) are discussed. As an organic dye for optical amplification, Rhodamine 6G, Rhodamine B, and Perylene Red are doped in the core region of polymer optical fibers (POFs). These POFA can obtain optical gain in the visible region of wavelength from 570 nm to 620 nm. POFA is promising for extraordinary high power optical amplification in comparison with rare-earth ions doped silica fiber amplifier. For example, output power of 1 kW with a gain of 30 dB can be obtained by using a Rhodamine B doped POFA at a low dye concentration of 1 ppm. Additionally, a novel solid-state POFA amplifier system is demonstrated.
Proc. SPIE 2289, Four by 622 Mb/s 40-km transmission experiment of wavelength-division multiplexing (WDM) system with erbium-doped fiber amplifier (EDFA), 0000 (3 October 1994); https://doi.org/10.1117/12.188706
A 4 X 622 Mb/s 40 km WDM optical communication system with EDFA is described. Four DFB-LD modules with 1.55 micrometers wavelength range are used in the transmitters. An EDFA with 1480 nm pump LD is used as a boost power amplifier. It suppresses noises efficiently. A Fabry-Perot interferometer is used as a demultiplexer. Through 40 km transmission, the output waveform of the receiver is obviously improved. It makes the research on long haul and WDM/FDM optical communication systems great progress.
A stable second-order nonlinearity established by temperature/electric-field poling of a SiO2-based waveguide stack on a Si substrate with a nonlinearity comparable to that observed in bulk fused silica samples (approximately 1 pm/V) is described. Samples with various layer thickness and composition are compared to determine important parameters in generating the second-order nonlinearity. Temperature and voltage studies are also presented to help understand the dynamics of the nonlinearity.
Optical second harmonic intensity has been measured for electrically poled several TeO2- based glasses in order to discuss relation between second harmonic intensity and glass structure. The glass systems, ZnO-TeO2, BaO-TeO2, WO3-TeO2 and ZnO- BaO-TeO2, have been chosen because these systems possess relatively extended glass- forming region. The second harmonic intensity increases with an increase in the concentration of ZnO in the binary ZnO-TeO2 system. It is known that an addition of ZnO to TeO2 glass converts TeO4 trigonal bipyramids to TeO3 trigonal pyramids and leads to creation of non-bridging oxygens. We presume that these tellurite glass structural units, which are lacking in center of symmetry, are oriented to the direction of external dc electric field and that the orientation takes place more easily in the ZnO-rich glass because the glass-network structure is more flexible owing to the presence of trigonal pyramids and non-bridging oxygens. While the second harmonic generation is clearly observed in ZnO-TeO2 and WO3-TeO2 glasses, it does not occur in BaO-TeO2 glasses under the present poling conditions. We speculate that the orientation of structural units scarcely occurs in the BaO-TeO2 glasses because the external field mainly causes the electron polarization of Ba2+ ions.
The level of second harmonic generation (SHG) in fused silica can be altered by the addition and removal of water by wet and dry annealing in N2. The magnitude of the effect is sensitive to the lot of material. Maximum SHG occurs with annealing temperatures near 500 degree(s)C with smaller values of SHG occurring for annealing temperatures on either side of 500 degree(s)C. For annealed samples, the SHG has been found to have both surface and bulk components. Preliminary EPR analysis has shown a growth in both the Si and Ge E' centers on either side of a 500 degree(s)C - 600 degree(s)C annealing temperature. Finally, from depoling current data, evidence for a second order `bond effect' along with a third order `hole filling effect' has been found.
Results are presented on the induction of large second order optical nonlinearities in high purity silica glasses by poling or a combination of poling and UV laser exposure for localization of the induced second order effects. Localization is required to achieve quasi phase matching in optical planar waveguides since the induced birefringence in these structures is too small to allow standard phase matching approaches to be used. Samples of Suprasil, Infrasil, and Homosil were poled using standard high temperature poling at 300 degree(s)C with an applied field of approximately 13 kV/cm. Correlation between the poling and depoling currents for these samples and the observed second harmonic generation is seen. The results support the thesis that the origin of the second order nonlinearity is the combination of a large surface field and the intrinsic third order nonlinearity in the glass. Relatively short room temperature UV (257 nm) laser exposure with beam power of a few W/cm2 was found to completely remove the induced second order nonlinearity in thermally poled glasses. This allows a simple fabrication technique for nonlinear or combined linear/nonlinear gratings in planar glass waveguides using thermal poling followed by localization by UV exposure.
Second harmonic generation is used to characterize the second order nonlinearity induced by thermal electric field poling in fused silica. The uniformity of the poled region is improved by using an evaporated rather than a machined poling electrode configuration. The effective thickness L and strength of the nonlinear coefficient d33 of the nonlinear region is estimated using Maker fringes excited through a prism pair.
We report the first dedicated study on the polarization properties of fiber lasers based on rare- earth-doped single-mode elliptical-core fibers. More specifically, with the use of elliptical-core (2:1) Nd-doped fibers, we report evidence of significant gain anisotropy, primarily due to the differences in the modal overlap in the two orthogonal polarizations.