The laser power was increased from 350mW to 1 W at the BS. The sensitivity was improved in all the frequencies. 

Although the high-frequency noise reduction makes sense, the low-frequency noise reduction is not really obvious.

 

1. IMC lock acquisition modification

 When we increased the laser power, the IMC got difficult to be locked by the guardian. It lost lock when the guardian engages the boost filters, and there was a huge offset on the feedback signal to the laser frequency. I modified the guardian so that the MCL loop is engaged during the lock acquisition, then the guardian could lock it. 

 We guess this is because the IMC mirrors were pushed by the radiation pressure when the IMC got locked. 

 

2. AS port laser power tweaking

 We rotated the AS HWP and adjusted AS PD so that the laser power on the PD is the same as before.

 

3. LSC gain normalization

 LSC gains were normalized. I used the INMON value of the single bounce from the ITMX. They are 360 (REFL PDA1), and 1600 (AS PDA1)

 

4. Sensitivity improvement.

 As shown in the attached figure, the sensitivity was improved at all frequencies. Above 100 Hz, we know that the REFL PD dark noise limited the DARM sensitivity, and it improved as we expected. 

 The low-frequency noises also improved, although the LF noise was not improved by changing the REFL PD power as shown in klog11775. It would be a hint of the noise source at low frequencies, but I cannot come up with good answer right now. One thing I checked is the MICH PD dark noise, but it is not limiting the MICH sensitivity as shown in attached figure.