[Kenta, Michimura (remotely)]
Using the s+p total power at POP and POS when the arms are locked and unlocked, Xarm and Yarm round-trip loss was measured to be the following:
Xarm loss from POS: 62 +/- 6 ppm
Yarm loss from POS: 74 +/- 6 ppm
Xarm loss from POP: 77 +/- 20 ppm
Yarm loss from POP: 42 +/- 23 ppm
Compared with the same measurement done at room temperature in Aug 2024 (klog #30823), the round-trip losses are ~10 ppm higher.
Measurement at POP has higher uncertainties due to lower gain in POP_SPOL PD.
Motivations:
- Since ITM reflectivity seems to be changing over time, especially at cryogenic temperatures (klog #36238), it is impossible to estimate the round-trip losses from the finesse measurements alone. Finesse roughly got back to pre-cool down after the warm up Dec 2025, but it does not necessarily mean that round-trip losses got back.
- We can estimate losses inside the PRC for sidebands with POP90, but it is hard to estimate the losses inside the PRC for carrier when the arm is locked. Carrier PRG gives the total of PRC losses and losses from the arm cavity.
- So, we wanted to see the latest round-trip loss after the warm up.
Method:
- We basically repeated the thing done in klog #30823.
- For ITM single bounce, data take on on January 29 was used (klog #36249).
- For Xarm and Yarm single locks, data taken today by Kenta was used. GR shutters were closed during this data taking. While K1:LSC-TR_X_IR_NORM_OUT_DQ was roughly 1.0, K1:LSC-TR_Y_IR_NORM_OUT_DQ was 0.9. ADS didn't make this 1 and the beamspots on ETMY looked OK.
- We zero-ed the offsets for {POP,POS}_{SPOL,PPOL}_DC and REFL_PDA1_DC before the measurements with shutters closed. For REFL_PDA1_DC, we forgot to do this for January 29 data. But this is not an issue since we are not using REFL for the analysis.
- For POP_SPOL, the data was multiplied by 2*10**(30/20) to take into account of extra BS in front of POP_SPOL PD and the PD gain difference (POP_SPOL PD has gain of 10 dB, while POP_PPOL, POS_SPOL and POS_PPOL PDs has 40 dB; see klog #36249). All the gains were 40 dB when we have done the measurments in klog #30823.
Results:
- Attachment #1 is the data from PDs. The numbers for K1:LSC-POP_SPOL_DC_OUT16 in the plot is multiplied by 2*10**(30/20), as explained above. REFL data seems bogus, probably because the offsets were not zero-ed for single bounce data.
- p-pol content calculated with p/(s+p) were the following (R:T=0.2:0.8 for BS for p-pol taken into account), see Attachment #2:
POS ITMX Single 5.64 +/- 0.02 %
POS Xarm Locked 0.71 +/- 0.03 %
POS ITMY Single 8.54 +/- 0.00 %
POS Yarm Locked 1.79 +/- 0.10 %
POP ITMX Single 6.79 +/- 0.10 %
POP Xarm Locked 1.32 +/- 0.11 %
POP ITMY Single 10.02 +/- 0.13 %
POP Yarm Locked 2.12 +/- 0.13 %
- RTL measured using the s+p total power were the ones stated at the top. Measurements done with POP has higher uncertainties since POP_SPOL PD has 30 dB less gain.
Discussions:
- p-pol ratio during the single bounce and single arm lock suggest that the Lawrence effect reduces the p-pol power by a factor of 5-8. Consistent with the measurements done in klog #30823, but larger than a factor of ~3 we observed initially back in 2019 (klog #9393).
- Having 5-6 % of p-pol from ITMX and 8-10% of p-pol from ITMY single bounce is consistent with ITM (effective) reflectivity ratio X/Y for s-pol of 1.039+/-0.005 and ITM p-pol conversion ratio X/Y of 0.663+/-0.005 (klog #36249).
- To explain observed PRG for PRX and PRY, losses in PRC were estimated to be ~10% for PRX and ~15% for PRY (klog #36249). These are ~2 times larger than the p-pol component. We might be having some extra ~5% losses other than birefringence in the PRC.
Next:
- Estimate PRC losses for carrier using recent PRFPMI data and finesse measurements.
- Re-calibrate K1:LSC-TR_T_IR_NORM_OUT_DQ so that it will be 1.0 for single Y arm lock
- We may want to implement the round-trip measurements in the finesse measurement script so that we can measure them occationally.
