PhotInd project (2015-2018)
The aim of this project was to develop characterisation and calibration techniques for the latest generation of photonic components and devices in both optical fibres and waveguides, which would underpin their development and manufacture. This would strengthen the competitiveness of the European photonics industry, enabling innovation and providing faster, cheaper data connections.
The contributions to WP1 included in the development of non-destructive and novel methods to measure dispersion and group velocities in optical fibres. The target was to measure optical dispersion with a relative uncertainty of 1×10^‑3 for dispersion slope and zero dispersion wavelength for 100 m of silica fibre.
The White Rabbit (WR) technique in its core principle is a time of flight measurement at different central wavelengths of the optical pulses. Our experimental setup of measuring the fibre dispersion is a custom made. The light from a supercontinuum laser source (L) in the spectral range of 1280 nm to 1680 nm is filtered with a MEMS-based electrically tunable Fabry-Perot interferometer (FPI). The light is coupled into a short fibre (F1) with an off-axis parabolic mirror. The fibre, with protective coating removed, is bent to the extent that a portion of the light travelling in the fibre radiates out through the cladding (bent loss). Most of the light remains in the fibre and will be guided to the fibre spool under the study (FS). After the fibre spool, the light is guided to a short fibre (F2), and a bent loss is introduced again. The light from the bent loss is focused on a fast optical detector (APD) and analysed to calculate the time the light was in the fibre.
We can estimate the dispersion based on the time the pulses with different central wavelengths stored in the fibre. Although in principle we could measure with this method the fibre with any length then in real situations the characteristics of the optical pulses sent into the fibre and the characteristics of the detector are limiting the minimum length of fibre that can be studied. Also for practical purposes, it is wise to use fibres that are characterisable with other methods to have directly comparable results.
We did a comparison measurement with the calibrated chromatic dispersion reference fibre (Ref. 4 G655 Leaf) that was developed by METAS. The length of the spool was approximately 10 km and based on this information we were able to estimate the chromatic dispersion of the fibre with our WR system. The measurements were done at the wavelength range from 1290 nm up to 1650 nm. The width of the optical pulses are presented in the figure below.
With WR system, delays of optical pulses were measured and a fourth order polynomial expression was fitted through the measurement points to describe the delay analytically. Based on the measurement and analytical fit the chromatic dispersion of the reference fibre was calculated. The results are presented in the figure below. The estimated relative uncertainty for that measurement of the chromatic dispersion is no more than 2% at the 95% of confidence interval. The zero dispersion wavelength is at 1498.8 nm.