Virgo Logbook

AdV-INJ (Input Power Control (IPC) system)

gosselin, melo - 23:45 Thursday 11 January 2024 (62937)

Swap the discarded power from IPC1 to the IPC on the LB

The goal of the shift was to swap the discarded power from IPC1 on EIB to the IPC before the AOM on the Laser bench in order to lower the power in the PMC, EOM and Faraday Isolator and see it had a positive effect on the IMC mode matching and on possible stray light from IPC1. The initial situation was indeed estimated to be 44 W at the input of IPC1, 20.5 W were transmitted (EIB_PD_OUT) and about 23.5 W discarded on the beam dump.

The first step was to set the amplitude of the line at 0.1 and check and adjust the working point of the IMC and the offsets on the RF QPD of IMC REFL since we were expecting this action to modify them.

IMC WP -0.08 -> -0.07

NF_ V -> -0.08 → -0.065 e mag a 5e-4
NF_H -> 0.14 -> 0.1 e mag 0.8e-4
FF_H -> 0.45 -> 0.17 e mag 5e-3
FF_V -> 0.4 -> 0.27 e mag 5e-4

Around 17:40 UTC We proceed with the swap of power.
IPC0 -> 0.8V (PMC_TRA) to 0.48
IPC1 ->recover 20.5W on EIB_Pout

We had to decrease the attenuation on the error signal of the PMC by 2 dB to take into account this power decrease.

There is now about 26 W at the input of IPC1, 20.5 W are still transmitted towards the IMC while 5.5 W are discarded in the beam dump. We kept some margin since IPC1 is used at the beginning of the lock acquisition to adjust the power in transmission of the IMC.

After having swapped the IPC we started to check the working point and offsets.

The values we found at the beginning were very high and far away from what we had before.
Then we realized that the BPC was actually not locked.
We investigated a bit and found out that there it was the set point of the trigger on PMC_TRA was till high and was preventing the lock . Paolo from remote helped us to fix it (first line of page 2 of BPCPSD)

So with the BPC relocked we worked again on the setpoints.

WP IMC 0.02 V mag a 15e-6

NF_V -> -0.02 e mag 1e-3
NF_H -> 0.06 mag 1e-3
FF_H-> 0.1 mag 1e-3
FF_V-> 0.5 mag 1e-3

After having set the new offsets of the galvo the mag of the monitor for the IMC WP increased from 15e-6 to 1e-4
We tuned it again and found 0.12 V with a mag back to 15e-6
Increase the line to 0.2 and check all the WPs, They looked ok.

Then we proceed with a manual tuning of Fmoderr to get the IMC in its nominal alignement situation and proceed with the sensing matrix for the automatic alignement.

Here below the UTC time for the noise injection on the various degrees of freedom:

Ib tx 20:44:58 2min
ib ty 20:47:15 2 min
Ib tz 20:49:25 2 min
MC tx 20:56:10 2 min
MC ty 20:58:30 2 min

The new alignment matrix is the following.

ACL_MATRIX_BEGIN   sensing_matrix   1   MCT_h_err   FF_ACh_err   NF_ACh_err       MCT_v_err   FF_ACv_err   NF_ACv_err
ACL_MATRIX_CH           MC_ty_err   ""       16.12       0.0     0.0             0           0           0
ACL_MATRIX_CH           IB_ty_err   ""       29.94       0.0     0.53            0           0           0
ACL_MATRIX_CH           MC_tx_err   ""       0           0       0               11.16        -0.69        0.42
ACL_MATRIX_CH           IB_tx_err   ""       0           0       0               42.72        -0.59        0.36
ACL_MATRIX_CH           IB_tz_err   ""       0           0       0                10.50       0.09        -0.9

It slightly changed compared to the previous one:

#ACL_MATRIX_BEGIN   sensing_matrix   1   MCT_h_err   FF_ACh_err   NF_ACh_err       MCT_v_err   FF_ACv_err   NF_ACv_err
#ACL_MATRIX_CH           MC_ty_err   ""       14.49       0.0     0.0             0           0           0
#ACL_MATRIX_CH           IB_ty_err   ""       28.00       0.0     0.83            0           0           0
#ACL_MATRIX_CH           MC_tx_err   ""       0           0       0               10.77        -1.07        0.67
#ACL_MATRIX_CH           IB_tx_err   ""       0           0       0               41.64        -0.88        0.55
#ACL_MATRIX_CH           IB_tz_err   ""       0           0       0                10.77       -0.25        -1.86

After having relocked all the loop of the injection we did check and fine tune again the WP before relocking the ITF.

WP IMC 0.02 V mag a 15e-6

NF_V -> -0.06 e mag 0.002
NF_H -> 0.04 mag 2e-3
FF_H-> 0.03 mag 2e-3
FF_V-> 0.5 mag 1e-3

We relocked the ITF without any trouble till LN3.

More analysis on the new mismatch of the IMC will follow.

Comments to this report:

melo, gosselin - 14:24 Friday 12 January 2024 (62942)

We have checked the PMC and IMC mismatch before and after the IPC power swap.

	Before IPC swap	After IPC swap
PMC Mismatch [%]	12.8	18.8
IMC Mismatch [%]	7	4.6

So, the PMC mismatch has worsened around 6%, while the mismatch towards the IMC has improved around 2.4%. The throughput losses on the Mode Cleaner have remained constant around 33%, calculating considering the power budget between the photodiodes INJ_IMC_TRA and INJ_EIB_POUT, as explained in #55656.

seglararro - 13:17 Monday 15 January 2024 (62971)

After this intervention, we have observed that the linear O3 noise subtraction that is done in a parallel Hrec thread (HrecDTF), devoted to cross-check the new noise subtraction (done in the main Hrec thread), is strongly affected by the 25- minutes glitch. Figure 1 shows some of the BNS range drops in HrecDTF and Figure 2 the effect on the TF. Figures 3 to 5 show the effect in the different Hrec. From the Omicron plots in VIM there are way more clusters at SNR<20 starting after this lock, and they exhibit different behaviour: they are way less regular in SNR and frequency. They are visible in DARM but also LSC_MICH and LSC_SRCL, used in the noise subtraction (Figures from 5 to 8).

Note that the noise subtraction in the main Hrec thread is not affected further as the TF used in this subtraction are set to zero in every cycle of 240s, whereas the noise subtraction in DTF is adaptative.

Images attached to this comment

derossi, gosselin - 15:59 Monday 22 January 2024 (63052)

Triggered by the plot of the IMC DC transmission on the DMS (plot 1), we investigated why the laser seems to have more intensity noise at low frequency.

The 2nd plot shows the INJ signals in time before (pink curve) and after (blue one) the swap of the discarded power from IPC 1 to IPC 0 (PMC DC in blue has been multiplied by 0.8/0.49 to take into account the action on IPC0). The excess of noise seems to be due to a misalignment of the PMC (the DC is noisier while the error signal is not). This had also been observed in laser lab when this operation was performed (with the IR card, looking at the reflection, the misalignment was visible by eye), and could be explained by the heating of the beam dump.

The 3rd plot shows the FFTs at the same gps than plot 2. We have more intensity noise below 2Hz, and it seems that on the photodiode EIB POUT there was a shoulder of scattered light which disappeared.

Images attached to this comment