It seems as if Polarization Maintaining fibers only exist in Singel Mode versions. Is that because the PM capabilities does not work for higher order modes, or has there simply never been any demand for such fibers?
If I launch polarized light into a PM 1550 fiber at a wavelength much below the cutoff, will I the light in the other end be (more) linearly polarized, or will the polarization be as "random" as if I launched the same light in a regular SMF28?
(sounds simple to just test but I don't have equipment to test this my self)
Thanks for your time/M
Dear Mikael Malmström ,
imagine a large core step index fiber in which the guiding of the light can be simply regarded by multiple reflections at the core to cladding boundary. Any polarisation state is destroyed during these reflections.
So to my opinion, polarization maintaining can only be achieved by single mode or only to very low modes; but in the latter case polarisation state is quickly destroyed by mode coupling.
There are some papers dealing with this subject, investigating elliptical core fibers.
Just 'google' for that subject.
Best regards
G.M.
Gerhard i dont understand why there shoud be a difference between SM and MM fibers?
If there is enough birefringence the modes of the MM fiber should be separeted enough so they can not transfer the energy between them? Or is it to hard to generate enough birefrinfenge in a larger core?
You are right;
the 'strength' of birefringence is limited as well as the dimensions of the elliptical core with respect to the wavelength.
So only for the low mode case the modes will 'preserve' their polarisation a while. However, as I have learned from some papers, coupling is still quite strong here, so that the polarization will go off quickly compared to the single mode in a polarization maintaining SM fiber.
I’ve tested the PM1550 and it actually keeps the polarization quite good at well below the cutoff. However, it is not a Multimode fiber but rather a Fewmodefiber so the output is not as smooth/filled as for a MM fiber but rather beating between different low order modes which is not what I wanted...
- Badges
- Science topic