CO2 laser intensity stabilization installation

Starting from monday 5th, we worked on the installation of the intensity stabilization loops on both TCS benches. Figure 1 shows the new layout of the tables. The writing "PD2" actually means the power meter. The polarizer before the wave-plate has been added to recover a good P polarization, since the AOM and the pick-up bs depolarize the transmitted beam. Two new cooling lines have been added for the AOMs.
The new components installed on the main beam shift the beam that is impinging the on the AXICON, so a re-alignment has been necessary. Figures from 2 to 5 show the thermal images respectively of the NI and WI rings before and after the re-alignment. It seems that the both rings have chaged a bit their position, to be checked once the ITF is relocked.
Because the Brewster attenuator is rotating the transmitted polarization, the reflectvity of the beam splitter is changed from the nominal value of 50% to maybe circa 66%. Thus the power impinging on the two PDs is not the same. It has been set such that the In-Loop PD is close to saturation, so the reading of the Ou-Of-Loop PD is limited by its own dark noise. Performances are anyway good (see below).
The electronics are the same on both benches:

• the In-Loop PDs are amplified with a low noise amplifier with a gain of 100
  
• the Out-Of-Loop PDs are amplified with a low noise amplifier with a gain of 300
  
• the signal of the In-Loop PDs is sent to a low noise PAR5113 pre-amplifier set in bandpass mode (input coupling AC, Gain=300, LF=30Hz, HF=300Hz)
  
• the output of the PAR5113 is fed to the RF driver of the AOM
  

We found a problem of grounding only on the NI bench, so we have to put floating both the PDs and the amplifiers. There was also a problem when connecting the output of the Out-Of-Loop PD to the ADC. It has been solved by using a modified BNC-Lemo adaptor: the negative is short-circuited to the ground.
The performances of the stabilization loop are good: see figures 6 and 7 for a comparison between the RIN in closed and open loop for the two benches. It is seems that the WI laser is particularly quiet: the free running noise is lower than the NI laser and also than the spare laser in Tor Vergata. Figure 8 shows a comparison of the RIN in closed loop measured in Tor Vergata Lab and on the Virgo site. The RIN is below 2e-7/sqrt(Hz) in both benches. It must be said that the NI laser is unstable and that the loop cannot keep the laser in the low noise state during these instabilities. Anyway it seems that the loop does not saturate during unstable periods. Work in in progress to better stabilize the working temperature of the NI laser.
Two new channels have been configured by A. Masserot (TCS_{NI,WI}_PstabCorr) to monitor the correction signal sent to the AOM. These channels have not yet been connected. Yesterday evening, we tried to connect the output of the PAR5113 to the ADC, but strangely the PAR was going in overload. This must be further checked.
The introduction of the new components required a re-calibration of the output power as a function of the ADC voltage and wave-plate angle Now the range of the two power meters has been set the same on both benches. Figures 9 and 10 show the TCS available power vs plate angle. Figures 11 and 12 show the TCS power vs ADC voltage. The reading of the NI power meter stops at about 10W, but actually the maximum available power is more than 12W. The same applies to the WI bench: maximum reading by the power meter is 9.5W, while the maximum available power is almost 10.5W.
The new calibration factors are:

• PNI(W)=5.526*TCS_NI_Power_DC - 0.475
  
• PWI(W)=4.864*TCS_WI_Power_DC - 0.195