Depolarizing and source linewidths

[lichen]

My question relates to fiber optics, and probably also general optics.

I have a broadband SLD source, which can be depolarized by a broadband fiber depolarizer. If I place a Fabry-Perot filter with low bandwidth between the source and depolarizer, I assume this reduces the linewidth going into the depolarizer and means the depolarizer has to be longer and more expensive to be able to handle this reduced linewidth of source.

What if a broadband depolarizer was used to depolarize the broadband source, and then the FP filter was placed after the depolarizer. Would the filtered light still be depolarized, or would this have the same effect as placing the filter before the broadband polarizer (i.e. the light would not be properly depolarized because the linewidth at the input is too low)?

Thanks.

 

[lichen]

Can anyone comment on this please?

The question is really, does placing an optical filter AFTER the depolariser, mean the depolariser doesn't have to handle such a low linewidth input?

 

[Andy Resnick]

Some additional detail would help- can you provide more information about fiber depolarizers? a manufacturer website would be fine. The only depolarizers I know of are things like rotating ground glass plates, which is bandwidth-independent (for all practical purposes).

As general comments go, I think FP devices are polarizing, because they work by interference.

 

[lichen]

Thanks for replying Andy.

The depolarizer I had in mind was http://www.phoenix-photonics.com/products/polarizers_depolarizers/Depolarizers_V12_1102.shtml by Phoenix Photonics, since it operates over 1300nm and 1550nm regimes.

They say the device can depolarize down to 0.1nm linewidth. If I place my scanning filter (0.04nm bandwidth) before the depolarizer, clearly it will not be able to depolarise the light properly due to the long coherence length of the filtered light.

However, I would like to know if it would be possible to use the following sequence: Source (>0.1 nm) -> depolarizer -> filter.

The problem may be that by filtering light to 0.04nm linewidth after the depolariser, I am still somehow reducing the linewidth of the chain down to <0.1nm. This is what I am unsure of.

 

[Andy Resnick]

From what I know, Lyot depolarizers are only useful for broadband light. They are static devices with a (spectrally) periodic change in output polarization state- the depolarizing aspect comes from superposition over wavelength.

http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=01072136

My suspicion is that by spectrally filtering the light prior to the depolarizer, you decrease the efficiency of depolarization, although I can't say by how much (perhaps the manufacturer can). I don't exactly follow your last sentence- what are you ultimately trying to accomplish?

 

[lichen]

Basically I am asking, what if I spectrally filtered after the depolariser instead of before?

The answer may be the apparently obvious: that everything would work fine.

But the answer may also be: that is the same as filtering before the depolariser, and your filtered light will ultimately not be depolarised properly.

If it is the latter, I cannot use the 0.1 nm depolariser in my system.

 

[Andy Resnick]

It's an interesting problem, because you are balancing two competing effects- a narrow bandwidth (which is a high temporal coherence) and a low state of polarization (which requires low temporal coherence). Just to use round numbers, 0.05 nm bandwidth centered at 1500 nm gives a coherence length of 4.5 cm... not as coherent as I thought, actually.

I would put the filter after the depolarizer. Can you measure the Stokes vector?