[Sheila, Elenna]
Continuing the work from alog 62833, I made an independent calculation of the power on the ASC REFL A and B QPDs by calculating the power reflected from the PRM.
PRM reflection:
The PRM is 97% reflective, and the IFI transmission is 0.977 (alog 46386 and https://dcc.ligo.org/DocDB/0119/P1500076/003/Mueller16rsi-AdvIO3.pdf). This beam passes through several HAM1 optics to the ASC REFL WFS (layout here: https://dcc.ligo.org/DocDB/0009/D1000313/014/D1000313-v12.PDF), specifically M1 (BS 50/50), M14 (BS90/10 R), M17 (BS 50/50), M6 (BS 50/50), and a final BS 50/50 to for REFL A and B (.5*.5*.5*.5*.1=0.00625). Craig notes in alog 62440 that the IM4 trans is underestimated by 9.5%. I also took into account that the IMC transmits 80% of the input power from the PSL.
**** Daniel has brought to my attention that the HAM1 diagram is missing an additional 50/50 beamsplitter that splits the light to in-air versus in-vacuum. That means that there is an additional 50% reduction. Editing below-
REFL DC NSUM:
See attached photo for estimated counts at each input power. The counts are estimated from the REFL_A/B_DC_NSUM_OUT16 channel. To calculate the power, I multiple by the ADC conversion 6.1e-4 V/ct, transimpedance of 0.001 A/V, and PD responsivity of 0.8 A/W.
| Input Power (W) | Power calculated from PRM reflection (mW) | Power calculated from REFL DC NSUM (mW) |
| 2 | 5.2 | 6.3 |
| 4 | 10.4 | 12.2 |
| 6 | 15.5 | 18.3 |
| 8 | 20.7 | 24.4 |
| 10 | 25.9 | 30.5 |
The values from the DC NSUM are consistently 60% of the power estimated from the PRM reflection.
The values calculated from PRM reflection are slightly less than the measured power from the NSUM calculation.
As confirmed with Keita, and also by Georgia's photo in the comment to this alog (63001), there are only 4 50/50 beamsplitters along the path to the ASC REFL WFS. Similarly, there are four 50/50 beamsplitters along the path to the LSC REFL WFS. This means that my original calculations of the estimated power from the PRM reflection were correct. As a further confirmation, the DC power on the LSC WFS is similar to the DC power on the ASC WFS. See table.
| Input Power from PSL (W) | PRM Reflection estimate (mW) | Power calculated from ASC REFL DC NSUM (mW) | Power measured by LSC REFL A/B LF (mW) |
| 2 | 10.4 | 6.3 | 7.3/6.8 |
| 4 | 20.8 | 12.2 | 14.7/13.5 |
| 6 | 31.1 | 18.3 | 22/20.3 |
| 8 | 41.5 | 24.4 | 29.5/27 |
| 10 | 51.9 | 30.5 | 37/34 |
PRM Reflection estimate: IM4 trans is 80% of the input power from the PSL, with an additional 9.5% correction based on the difference between Keita's measurement and Craig's calibration. The PRM is 97% reflective, the IFI transmits .977 of the power, and one 90/10 beamsplitter and 4 50/50 beamsplitters give the expected power on the REFL WFS (ASC and LSC).
ASC REFL DC calibration: +-20 V 16 bit ADC, transimpedance of 1 kOhm, 0.8 A/W PD responsivity
LSC REFL LF calibration: +-20 bit ADC, transimpedance of 200 Ohms, 0.8 A/W PD responsivity
The values on the WFS are approximately 60% less than the values from the PRM estimate.