SRS is reduced by increasing the bandwidth of the OC in high power fiber lasers generally. What happens if the bandwidth is increased too much?

Technical Discussions
1

Several articles suggest, in a MOPA configuration high power fiber laser, the SRS can be mitigated by increasing the bandwidth of the OC FBG. How to determine the optimized bandwidth of such FBG? Does the bandwidth of the HR FBG is also related? What is the relationship between the bandwidths of the FBGs and the amount of Raman output?

For example, we saw an SRS peak by using a 1.6nm OC in a high power fiber laser. In order to reduce the SRS effect we inserted a 3nm bandwidth OC instead of the 1.6nm. Instead of a reduction in Raman, we observed multiple Raman peaks with the 3nm with a lot of back reflection. Our HR FBG is 5nm.

2

Thanks for the question.
Please could you share the links to the articles?

In your case, it is clear that increasing the OC FBG bandwidth to mitigate SRS had unintended consequences both in increased number of back reflections and number of raman peaks. This could be due to the interaction of the broader spectrum of the gain medium and the fiber itself. I.e increasing the bandwidth increases the number of modes that could contribute to the observed Raman peaks. Whilst I am not an expert in high power fiber lasers systems, I would suggest an iterative approach to fine tune your system until the best tradeoff is reached. The approach should experiment with both narrower bandwidth testing (between 1.6 and 3 nm) and HR FBG optimization.

The bandwidth of the HR FBG can also affect the gain profile and the distribution of energy within the cavity. For example it could be that the narrower (5 mm) bandwidth may be emphasizing certain gain regions that interact more strongly with the broader bandwidth of the OC FBG. This is why I would recommend optimizing both the OC and HR FBG bandwidths together to reach an optimum.

3

SRS is a tricky topic. It’s Stimulated Raman Scattering, just like Stimulated Emission. Basically if you have enough seed photons and SRS gain then it works as a laser.
Now, assuming the SRS gain is constant, by increasing the bandwidth of the radiation you are decreasing the number of seed photons per nanometers. So double the bandwidth should get the SRS threshold to double the power (photons per seconds).

There are simulation tools for that if you want to have a more accurate answer for your whole system. VPI Photonics and RP Photonics both offer excellent simulation tools for this. Alternatively you could write your own Matlab code for this too.

Best of luck
Daniel