Polish both ends. Are you sure your light is fully focused into the core?
to reduce optical fiber losses, Rayleigh scattering loss, infrared absorption loss, and losses due to imperfections should be reduced. Among these, Rayleigh scattering and infrared absorption losses are highly affected by used dopant. On the other hand, losses due to imperfections depend not only on the fabrication process but also on the viscosity properties of the core and cladding materials. To reduce loss due to imperfections, the viscosity matching of core and cladding materials is quite effective.
T
How is the coupling done into your fiber? A lot of loses are due to coupling losses, polishing both ends and improving your coupling efficiency by optimizing the focussing on both ends should help.
The replies above suggest you need to provide more details if you want useful advice.
What kind of fibre are you using, or trying to fabricate? What is the operating wavelength? Is it single mode or multimode? What is the fibre length? Do you need high bandwidth for data transmission, or high numerical aperture to couple light from a divergent source?
You can buy single mode silica core fibre with attenuation in the range 0.15 to 0.17 dB/km at 1550 nm.
Polymer optical fibre is readily available with attenuation around 200 dB/km at 650 nm, and losses of 50 dB/km are possible, depending on the core composition and operating wavelength.
For visible and near IR multimode fibres, flame hydrolised ("high OH") fibre is preferred for shorter wavelength visible and UV operation. "Low OH" fibre has lower absorption loss at near infra-red wavelengths..
As Adam and Michael point out, input and output coupling loss are important, especially for short links.
Bending and microbending losses can be significant in over-filled multimode fibre, and some kinds of single mode fibre. You can measure anomalously high losses if you operate single mode fibres close to cut-off and excite higher order modes.
http://global-sei.com/news/press/13/prs032_s.html
http://www.corning.com/opticalfiber/products/SMF-28_ULL_fiber.aspx
http://uk.rs-online.com/web/p/fibre-optic-cable/8019279/
http://www.thorlabs.de/newgrouppage9.cfm?objectgroup_id=6838
Are you making your own fibers and have technical capabilities for doing it?? Then, you should have already expertise in place. Otherwise, your question is very vague and unclear.
Hi .. i need to reduce the losses results from extrinsic absorption bands due to
impurities, such as OH vibrations associated with water vapor dissolved in the glass and metallic-ion impurities in multimode fibers with wavelength 1.4 um
We are still short of relevant information.
What is the glass composition and fibre structure? Is it all-glass or polymer clad?
How are you making the glass? I assume you use some kind of vapour deposition process. This generally gives better control of impurities than double crucible or other multicomponent glass process.
What is the attenuation at 1400 nm, or rather at the hydroxyl absorption peak? Have you got hundreds or thousands of dB/km, a few dB/km, or tenths of a dB/km? If the loss is too high to measure, what is the absorption near the 1250 nm or 945 nm peaks where OH attenuation is lower? What is the attenuation at wavelengths far from the hydroxyl bands? Are you able to estimate the contribution from other impurities?
Do you use an inside tube process such as MCVD, or soot deposition followed by consolidation as in VAD or OVD? Do you use thermal oxidation or plasma deposition?
What source compounds do you use? Liquid or gas? Very high purity Silicon and Germanium compounds such as SiCl4 and GeCl4 are made for the semiconductor industries, but may not be optimum for optical fibre manufacture. Optical fibres are particularly sensitivity to trace quantities of transition metals, but optical absorption is less sensitive to alkali metal impurities which would cause major problems in semiconductor fabrication..
It is important to keep the vapour transport system clean and leak-free to avoid ingress of moisture and subsequent problems with corrosion. Welded stainless steel is less likely to leak than glass and PTFE pipework, but may be harder to diagnose and fix if it does leak. One option to minimize water ingress is to enclose the pipework and joints and flush the space around them with dry nitrogen.
Water vapour dissolved in the glass is unlikely to be the source of your hydroxyl absorption. The absorption arises from Si-O-H chemically bonded into the glass. If you use a dry process with thermal oxidation of chloride vapours, chlorine released during oxidation will react with free water and remove most of any hydrogen as hydrogen chloride. A more problematic source is hydrogen chemically bonded to silicon in the source material, so it is important to use source materials with low concentrations of such impurities.
If you use a soot consolidation process, thermal treatment of the porous soot with chlorine may be effective in reducing hydroxyl contamination.
The loss of optical fiber had been reduced to the point where the main contribution is Rayleigh scattering loss. At that point we thought we reached the limit. Actually it turns out that the Rayleigh scattering loss in a fiber is due to minute index fluctuation thermally excited but frozen when the glass solidified around 900-deg C or so.
I think the latest low-loss fiber uses a dopant to lower this glass transition temperature in order to reduce the Rayleigh scattering loss.
Is it single mode or multi-mode fiber?
The appropriate combiners may optimize the coupling losses. Short radius of bending may become significant in multi mode fibers.
PCF may contribute as a low loss single fiber for short distance applications.
fine modulation of the laser beam used with this fiber can reasonably reduce the losses of this fiber. As well, keeping the fiber at relatively constant temperatures may assist to do so
Parviz:
Layth is using multimode fibre, and is concerned about intrinsic fibre absorption rather than coupling loss.
He has not told us whether a step index profile is acceptable or whether a graded index profile is needed to reduce modal dispersion.
Oday:
Frequency modulation of the source can suppress modal noise, but will not help with the transition metal and hydroxyl absorption losses that Layth wishes to reduce.
Need more data: what is the experiment you are planning?
Simply speaking:
avoid curves or use bigger radii of curvatures
Use MM if SM are of no importance.
Use shorter wavelength if possible.
Select fiber models with better background losses if you have such option.
Again - please provide more data...
Yariv,
I agree we need more data, but some of your questions have already been answered - with my browser we are on the second page of answers.
Layth is using multimode fibre and is concerned about absorption from OH and transition metal impurity. We don't know what levels he is concerned about.
For reasons he has not disclosed he is operating at 1400 nm.
Commercial fibres rarely show transition metal contamination these days, so it would be good to know what fabrication process he is using. Is the OH concentration above 100 ppm, or less than 1 ppm?
If the intrisic OH or transition metal impurities are the sources of concern, then low OH concentration and highly pure fibers should be used. There are several commercial fibers with low OH-core 1-20 ppm, particularly at NIR spectral range. I have no comment for OH trace below 1ppm.
- Badges
- Science topic