Dear Abbas,

I suggest to you a links and attached files in topics.

- Self-heterodyne Linewidth Measurement - RP Photonics

https://www.rp-photonics.com/self_heterodyne_linewidth_measure...

- Laser linewidth measurement - Koheron

https://www.koheron.com/blog/.../laser-linewidth-measurement.ht...

- Delayed Self-Heterodyne Phase Noise Measurements ... - IEEE Xplore

ieeexplore.ieee.org/iel5/68/6133462/06064871.pdf

- OSA | On the delayed self-heterodyne interferometric technique for ...

https://www.osapublishing.org/oe/abstract.cfm?uri=oe-14-9

Best regards

Hi Abbas,

The linewidth of a laser is intrinsically related to the spectral density of frequency noise. So when you change the span of your ESA, you basically change the time window over which you measure your linewidth. A longer integration time means that you integrates more of the low frequency fluctuation of the laser and the measured linewidth gets wider.

Here is an excellent paper (open source) describing how to infer the linewidth (and more) from a self-heterodyne measurement.

I hope this helps.

https://www.osapublishing.org/oe/abstract.cfm?uri=oe-24-24-27961

Dear Alexandre Laurain , Mohamed-Mourad Lafifi, Thank you so much for your invaluable suggestions.

Dear Alexandre Laurain, You might forgot to write the link in your comment.Could you plz provide the link? thanks

https://doi.org/10.1364/OE.24.027961

Can I check that you have an optical isolator between the laser and the fibre interferometer you are using in your measurement? If not, there is a good chance that the laser spectrum will be significantly modified by feedback due to coherent raleigh backscatter from the first part of the 20km reference fibre.

yes I am using isolator after DFB_LD,Please see the attached picture.Instead of 7Km ,I am using 20Km fiber.

Hi Abbas,

I am not sure if use of PM is correct. According to the literature and my experience, AOM or AOFS (frequency shifter) should work.

5 MHZ should be easy to catch. The rule of thumb is to use the delay line of 4 to 6 coherence lengths, which would be just around 100 m in your case. With 20 km you are measuring white noise rather than the linewidth.

Lastly, jitter will also contribute to your linewidth measurements. It will be more pronounced with more averaging.

Best of luck!

Dear Anton Khomenko, Thank you so much for valuable suggestions.I am quite sure that I am missing the things you've mentioned.I will follow your suggestions,thanks

Dear Anton Khomenko, can you tell me what is the purpose of using AOM instead of PM? thanks

Hi Abbas,

AOM shifts the signal frequency from 0 to the frequency of the AOM/AOFS, so you can measure the linewidth correctly.

Just make sure to use the detector and spectrum analyzer with suitable bandwidths.