The first part of the shift was dedicated to aligning the carrier pick-off polarization to the fast axis of the fiber that goes from injection to atrium

Polarization State Measurements of INJ_FIBER and BOX_FIBER:

The polarization state at the output of the INJ_FIBER and the BOX_FIBER was characterized (see Fig. 1). To avoid exceeding the maximum input power of the polarimeter (10 mW), the optical power was attenuated by adjusting the half-wave plate placed before the PBS. The resulting output powers were:
INJ_FIBER: ~6.4 mW
BOX_FIBER: ~3.2 mW

Measurements were performed using a Thorlabs PAX1000IR2 polarimeter. The fiber outputs were coupled to a collimator, which was subsequently aligned to the polarimeter photodiode through the instrument’s built-in alignment procedure.
A preliminary polarization optimization was carried out by adjusting the half-wave plate located after the PBS. To further minimize the ellipticity of the polarization state and improve the degree of linear polarization, a quarter-wave plate was then inserted downstream of the half-wave plate.

Extinction ratio (ER) measurements were conducted by thermally stressing the fibers, either manually or by using a hot-air gun at ~80 °C, to quantitatively evaluate the degree of linear polarization of the transmitted light. It should be noted that the ER values were not corrected for central ellipticity, as is typically done by a dedicated ER meter. The definition of this quantity is provided in the Thorlabs PAX1000IR2 polarimeter manual (pp. 70–72), available on the Thorlabs website.

According to manufacturer datasheets, typical polarization-maintaining (PM) fibers provide an ER of at least 23–25 dB. For the INJ_FIBER, the measured ER was ~32 dB, corresponding to a degree of polarization of 99.5%. Since 32 dB > 25 dB, this indicates proper polarization alignment. See Fig. 2.

For the BOX_FIBER, the measured ER was ~12 dB, corresponding to a degree of polarization of 90%. Since 12 dB < 25 dB, this indicates poor polarization alignment. See Fig. 3. This values was close to the initial one, indicating that the polarization degradation happens after the INJ_FIBER.
It should be emphasized that the BOX_FIBER is located downstream of several fiber splitters between the INJ_FIBER and the BOX_FIBER. This configuration suggests that polarization misalignment or degradation occurs along the optical path, which warrants further investigation. The latter optimization will be done in future shifts.

After the tests the lambda/4  installed between the lambda/2  and the fiber collimator is left in the injection line. Moreover after test we  restablished the initial power, increasing it a little bit (~15-20%).

After a preliminary analysis, it doesn't seem that the measured phase noise is improved by this procedure (probably due to the fact that the BOX_FIBER has still too low ER). See Fig. 4.

Finally , we've observed a power oscillation of about 5% of INJ_FIBER power output (and 10% of BOX_FIBER). This is something that we need to investigate better in the next shifts.

In the second part of the shift, we closed the phase noise cancellation loop, observing the effect on the lock accuracy of the longitudinal control of the FC. Below are the GPS times and the loop gains tested.  Each configuration is acquired for at least 120 seconds. 

The last point is taken by closing the loop by acting on the frequency of the main laser as well to adapt it to that of the cavity ("double loop") .

Data analysis of the different configurations will follow in short.

At the end  we opened the loop and left the system as it was before the shift.