Got enough squeezing to move on to injecting it into the IFO I think. Plot below shows sqz/asqz data points from H1 locking scheme (o) vs. L1 locking scheme (x). 20dB attenuator was in during H1 scheme measurement and 10dB attenuator was in during L1 scheme measurement. Similar CLF power were used (~1uW transmitted). The nlg calculation has taken dark noise into account.

Funny thing I observed was that if I increase CLF input power by a factor of 6 where I know I should have cleared the sensing noise, while HD didn't saturate I couldn't get more than 2.5 dB of squeezing at nlg of 2.285 (asqz 5.86). I tried increasing CLF and LO gain and non of them helped. 

Noise budget coming up next.