I have made a noise budget for the AS A WFS, RF36 and RF45. RF45 is used in full lock as the DHARD sensor and RF36 is used as the MICH ASC sensor. I have made similar budgets in the past for the AS WFS, but they contained some errors which caused confusion. For example, I was missing an additional 20 dB whitening chassis gain that confused my shot noise calculation.

Locked spectra and power were taken from the previous weekend's long lock, using GPS time 1357137858. I took dark noise data (with and without whitening) today during maintenance with the IMC offline.

See attached AS WFS budget plot for both RF45 Q and RF36 Q.

We are shot noise limited in AS A RF45 Q, and dark noise limited in AS A RF36 above 10 Hz. The Pcal lines between 10-20 Hz are evident in RF45, and therefore DHARD. Currently, DHARD P contributes to DARM the strongest from 20-30 Hz and around the calibration lines (alog 66698).

Calculation details:

DC calibration:

In lock, AS_A_DC has about 64,000 cts, or 4.8 mW of power. This is using 16 bit 20V ADC, transimpedance 1000 Ohms, 1.25 W/A, and 20 dB chassis whitening gain.

ASC_AS_C shows 0.72 W at this time (this is watts into HAM6 calibrated by Keita). OM1 and 2 transmission is 800 ppm, OM3 reflectivity is 0.99. Behind OM3 are two HR mirrors (M12 and 13 on HAM6 diagram), one lens (L101), and a 50% beamsplitter before AS_A (M101) (see D1000342, E1000457). Using those values there should be about 3.55 mW of power on AS_A_DC.

At this time, AS_B_DC has 68,000 cts, or 5.1 mW (using same W/ct above). There is slightly more power on AS_A/AS_B than predicted.

RF calibration:

RF45 uses 3 dB of whitening, and RF36 uses 12 dB. There is an additional 22 dB gain factor for RF input to Q output (see DCC T000044). In lock RF45 has another 1.995 ct/ct gain (unlocked 1 ct/ct) and RF36 has 5.587 ct/ct (unlocked 2.8 ct/ct). RF45 transimpedance is 800 Ohm, RF36 is 900 Ohms.

ADC/Dark estimation:

After taking data with the IMC offline, I turned off the stages of whitening and antiwhitening (3 stages RF45, 1 stage RF36) and took data to estimate the ADC contribution.

nowhit^2 - dark^2 = n_adc^2(1 - AW^2)    AW = antiwhitening filters, nowhit=no whitening filters applied

"nowhit" = n_adc + n_electronics, "dark" = n_electronics + n_adc*(AW)^k, where k is the number of whit/AW stages used