Elenna, Sheila
In alog 62219, I took spectra of the AS RF45, AS RF36, REFL RF45, and REFL RF9 A and B wavefront sensors with the whitening and anti-whitening filters turned off. I compared those measurements to reference plots of "dark noise", in other words, a combination of electronics and ADC noises. With Sheila's help, I learned how to estimate the ADC noise alone based on these two measurements. The plots I have attached to this log show the reference (all dark noises) compared to my calculated n_adc for the PIT and YAW channels of each WFS.
My calculations:
The reference noise measurement is a combination of electronics noise (n_e) and ADC noise (n_adc), multiplied by the corresponding whitening (W) and anti-whitening (AW) filters
ref = sqrt(n_e^2(W*AW)^2 + n_adc^2(AW)^2)
In very simple terms, the effects of the whitening and anti-whitening on the electronics noise should "cancel out" (W*AW=1).
ref = sqrt(n_e^2 + n_adc^2(AW)^2)
The measurements I took without any whitening/anti-whitening applied is simply
meas = sqrt(n_e^2 + n_adc^2)
Therefore I can calculate n_adc as
n_adc = sqrt(meas^2 - ref^2 / (1-AW^2))
When I plotted the ADC noise, I reapplied the anti-whitening filters so I could see the full effect of n_adc on the ref signal. The anti-whitening filters have a zero at 10 Hz, pole at 1 Hz and gain of 0.1. Each WFS channel had different numbers of whitening/anti-whitening filters applied.
AS A/B RF45: 3 filters
AS A/B RF36: 1 filter
REFL A/B RF45: 2 filters
REFL A/B RF9: 2 filters
I understand these plots to show that we are not ADC noise-limited in the AS and REFL WFS. Only AS RF36 looks worrisome, but I believe that is because only 1 anti-whitening filter is applied to the ADC noise.