craig.cahillane@LIGO.ORG - posted 18:03, Monday 16 May 2022 - last comment - 15:16, Thursday 19 May 2022(63154)
LSC POP path photodiodes during single bounce configurations
A couple weeks back, Elenna took some 50 W and 60 W OMC scans in both ITMX and ITMY single bounce config (alog 62883). I wanted to use this time to estimate the POP path efficiency, as we did for REFL some years ago (alog 46386). Livingston recently thought about this as well (LLO alog 59818) However, I'm not sure we can use this data to draw conclusions, because in these single bounce configs the POP path alignment is not controlled. Keita tells me that the LSC POP path is manually set using picos once the full interferometer is locked, which does not necessarily mean the path is well-aligned in single bounce. In the PNG attached, the first powerup is the ITMX single bounce, and the second is the ITMY single bounce. On the POP A diode, it is obvious that there are peaks and valleys on the order of 10% of the total power as the beam moves around and falls off the diode. The POPAIR diodes do not register sensible results. But, suppose we do some shoddy estimates with this beam. 1), we assume the peak power measured on POP A LF is when the beam was best aligned. This peak POP A LF power was 9.34 uW. 2), we estimate the "true" input power, recalling Keita's 9.5% calibration correction on IM4 (alog 62213). This "true" input power on the back of PRM was 53 W. 3), we calculate the input beam to POP A LF efficiency. This was done in https://git.ligo.org/aligo_commissioning/power_budget/-/blob/master/code/pop_path_calibration_via_single_bounce_config.py, and looks likePOP path power efficiency = T_PRM * R_PR2 * R_PR3 * T_BS * R_ITMX * T_BS * R_PR3 * T_PR2 * T_M12 POP path power efficiency ~ 1.7e-7whereT_PRM = 3.1 % T_PR2 = 229 ppm T_PR3 = 3.17 ppm T_BS = 50 % T_ITMX = 1.5 % T_M12 = 10 %M12 comes from a 90:10 BS called M12 in HAM1 that sends 90% of the POP light to the POPAIR path on ISCT1 (see HAM1 assembly DCC D1000313). With this POP efficiency and "true" input power, we estimate 9.27 uW should illuminate POP A LF. This is in near perfect agreement with the peak POP A LF measured of 9.34 uW. I think this lends some confidence to Keita's correction on our estimated input power, and suggests that our POP path has very little unexplained loss (unlike the REFL path). I'm putting together a reference with power budget stuff at https://git.ligo.org/aligo_commissioning/power_budget
Images attached to this report
Comments related to this report
The above result is perhaps in conflict with the result from alog 62936. The above measurement claims we were capturing *more* power on the POP LF PD, not less. alog 62936 estimates around 10% power is missing. It's possible that the maximum measured power on POP from the single bounce configuration time is not a very trustworthy measurement. Or maybe some losses come about on the POP path at higher power, or from the full power alignment of the POP path.