Today, after Michal noticed that the B1 beam was not properly centered on the B1 PD2 photodiode (logbook entry 44486), we investigated this issue, first by trying the recenter the beam with picomotors, by scanning the angular position of the SDB2 bench, and also by playing with the BS TY setpoint. The maximum power that could be reached on this photodiode was only a third of the expected level (that could be deduced from B1s2 when the OMC2 was out of resonance). Moreover during the angular scan of SDB2, we could not observe any plateau on this photodiode. This suggested that the beam was clipped somewhere (on the photodiode or on a mount).
Thus we decided to open the SDB2 minitower in order to inspect the bench for a possible clipping along the B1 beam path. Finally we noticed that one of the pawls of the B1_M2 mount that is normally used to hold the 2 inch optics was not properly positioned and did not seem in contact with the HR mirror. After repositioning correctly this pawl against the mirror, it was immediately possible to recover the expected level of power on the B1 PD2 photodiode.
The conclusion is that the miscentering of the beam on the B1 PD2 photodiode was due to a misalignment of the HR mirror which was not correctly mounted during last week intervention (logbook 44458). A lesson that we learn from this mistake is that such kind of operation (replacement of optics on a suspended bench) should be done with a beam available in order to have the chance to check for the correct alignment of the manipulated optics.
The minitower venting started around 13h00 utc. The pumping was started at about 15h20 utc (see Fig.1). Alessandro could restore the control of the suspensions towards 16h20 utc.
Unfortunately the torque of the suspension wire seems to have increased, as the correction signal in TY has increased by almost 1V (see Fig.2) after the suspension loops were closed. One can also notice that the bench moved by about 500 um in Z.
After the SDB2 intervention, Michal adjusted the relative alignment between OMC1 and OMC2 using the single bounce beam from NI aligned. During this operation both OMC cavities were locked (more details to come in a dedicated entry).
Tonight a period of Fast UnloCK crisis troubled the injection system.
We exploited this pathological state to take data with different configurations:
At around 17h41 LT, we switched the IMC rampoto' roll-off for the piezo actuation from 1/f^4 to 1/f. The IMC immediately relocked and did not lose lock for some ten minutes until we switched back to 1/f^4 at 17h48.
At 20h13 LT, triggered by a new crisis, again we switched to 1/f. Same outcome.
20h23 LT back to 1/f^4
For the record, the other actions we tried:
20h30 LT 1/f
20h40 LT 1/f^4
(unlock NOT in coincidence with the switch, some 30 seconds afterwards...)
at 21h12 LT
IMC gain dB lock range: 10 (locks also at 7, but from refl image there is a fast oscillation) to 22
at 21h14 set at 16 (was 15)
at 21h26 set back at 15dB
22h42 LT 1/f
(trying to lock arms - fast unlocks - set ITF_MAIN down)
(INJ_MAIN stuck in RFC_SCAN, ignores redirect)
(find RFC resonance by its own, then again FMODERR via MC_Z)
After recovering IMC_Restored, we tried again to lock the ITF, this time the IMC was much more stable. Lock acquisition attempts of the ITF were killed by other issues (see entry).
The rampoto' is left with 1/f configuration for the night. We will analyze data for the different periods to spot why with the 1/f the lock of the IMC seems much more robust even while trying to lock the ITF. This seems anyway to point towards something related to the EOM actuation.
Side note: no attempts to open the PStab loop or the alignment were made.
Tonight we tried again the ITF recovery, but we experienced lots of IMC instabilities (this will be described in a dedicated entry).
The few times that the IMC was quiet and collaborative, we tried to go to LN1, but we could reach this state only once over 3 attempts, and the other two times we got stuck at MICH_SET_0_1 with the warning: "WAITING FOR PR ALIGNMENT TO BE STABLE".
Looking at the signals, there is something strange in the corrections of the PR (see attached figures). It seems that they are active for few seconds and then they go back to zero.
To be investigated by the experts.
Upon arrival, I found ITF not locked and the activity on SDB2 in progress.
The problem on SDB2 has been fixed at 16:11 UTC. Then the evacuation of the SDB2 minitower started. Concluded at 18:00 UTC.
The rest of the shift was dedicated to the recovery of ITF. It has been done With the help of Chiummo for Injection system, Bersanetti, Gouaty, and Allocca.
ITF very unstable, Inj OnCall intervention was needed but problem still present.
The planned activity of OptChar / flat noise investigation has been postponed.
VPM
MdVim crashed several times. restarted properly
USGS crashed several times. restarted properly
16:40 UTC PyHvac restarted
17:03 UTC TFMoni crashed / restarted
INF - ACS
at 15:41utc I set the detection lab UTA in "portata ridotta" (SDB2 activity concluded)
This afternoon, thanks to the 1 MHz channel PSTAB_HF_CORR that Alain just added, we were able to compare the glitches occurring just before fast unlocks (already mentioned in 44457).
We took the IMC unlock which occurred at 15:24:51 UTC, and zoomed on the first glitch, shown on figure 1.
Figure 2 shows 3 channels, all with 1 MHz sampling:
- INJ_EOM_CORR is the correction sent from the IMC rampeauto to the EOM (the 14 MHz on the LB, just before the slave laser)
- INJ_IMC_REFL_I_PRE is the IMC error signal (reflection from the IMC, demodulated at 22 MHz)
- PSTAB-HF_CORR is the correction sent from the PSTAB rampeauto, whose photodiode error signal is located on the lower part of SIB1 and monitors a pick off of the trasmitted beam. The correction is sent to the AOM just before the PMC.
From the 2nd plot seems quite clear that the glitch occurs at first on the EOM CORR signal (blue line), and we found it with 6 microseconds delay on the IMC error signal (orange line) and finally on the PSTAB correction (green line)
-inf
at 13:00utc I set the detection lab UTA in "portata nominale" (SDB2 minitower opening)
some plots showing the microphone and DARM noise for 1minute around the times reported in the table for the helicopter flights over CEB:
- figure 1 - spectrograms, helicopter above CEB at altitude 2000ft
- figure 2 - spectra, helicopter above CEB at altitude 2000ft (compare qui quiet time)
- figure 3 - spectrograms, helicopter above CEB at altitude 1500ft (compare qui quiet time)
- figure 4 - spectra, helicopter above CEB at altitude 1500ft (compare qui quiet time)
- figure 5 - spectrograms, helicopter above CEB at altitude 1000ft (compare qui quiet time)
- figure 6 - spectra, helicopter above CEB at altitude 1000ft (compare qui quiet time)
ENV EERoom DAC channel is now: DAQbox 40, mezzanine 26, channel 4
First changed the mixing of B1s2 and B1_PD2 for the OMC1 error signal (the SDB1_B1x channels), as the ratio of power has now changed from 1/6 to 1/12.
This didn't have an impact.
Then noticed that the when OMC2 goes through the resonance the total power between B1s2 and B1 is not kept constant.
This is because SDB2_B1_DC has been for the past year the average of the two B1 PD powers, and with the current balance weight of 0.0 for B1_PD1 and 1.0 for B1_PD2 it actually corresponds to the total power on the B1 beam (so double the power in the situation with the 50/50 beam splitter).
Tried to divide the darm_b1dc_weight by a factor 2, it kept the total B1s2 and B1 power constant, but the ITF still unlocks.
To make it a bit cleaner (in case SDB2_B1_DC is used somewhere else), changed the value of the SDB2_ONE constant from 1.0 to 0.5, in this way SDB2_B1_DC = 0*B1_PD1 + 0.5*B1_PD2.
This didn't solve the problem either.
Figure 1 summarizes the situation when OMC2 goes through the resonance. The total B1s2 and B1 power is constant 12*B1s2 + B1_PD2 = 18mW.
But the B1p_56MHz_I signal is clearly moving away from zero as if the dark fringe offset is increasing.
An explanation could be that the beam is not well aligned on B1_PD1, and that only ~20% of the light reaches the photodiode.
Tonight's shift was dedicated to the ITF recovery.
We closed the PR angular loop in full bandwith in LN1, while the BS angular loop is only closed in drift control (put "1" instead of "2" in the BS.ramp_attribute("AATX(Y)_FLAG") in the ITF_LOCK.py) and we also slightly retuned the offsets.
The OMC1 lock didn't give much problems, but it was not possible to achieve the lock of the second OMC. We think it could be related to DARM handoff from B1s2 to B1: maybe, besides the "b1_pd_balance" there is some other value to be changed accordingly, but we are not sure.
SIDE notes: during the shift the ITF unlocked many times because of:
- fast unlocks
- PR misalignment
- earthquake, which excited the suspensions for about one hour
Since we couldn't reach LN3 we couldn't perform the scheduled negative DARM offset scan.
First changed the mixing of B1s2 and B1_PD2 for the OMC1 error signal (the SDB1_B1x channels), as the ratio of power has now changed from 1/6 to 1/12.
This didn't have an impact.
Then noticed that the when OMC2 goes through the resonance the total power between B1s2 and B1 is not kept constant.
This is because SDB2_B1_DC has been for the past year the average of the two B1 PD powers, and with the current balance weight of 0.0 for B1_PD1 and 1.0 for B1_PD2 it actually corresponds to the total power on the B1 beam (so double the power in the situation with the 50/50 beam splitter).
Tried to divide the darm_b1dc_weight by a factor 2, it kept the total B1s2 and B1 power constant, but the ITF still unlocks.
To make it a bit cleaner (in case SDB2_B1_DC is used somewhere else), changed the value of the SDB2_ONE constant from 1.0 to 0.5, in this way SDB2_B1_DC = 0*B1_PD1 + 0.5*B1_PD2.
This didn't solve the problem either.
Figure 1 summarizes the situation when OMC2 goes through the resonance. The total B1s2 and B1 power is constant 12*B1s2 + B1_PD2 = 18mW.
But the B1p_56MHz_I signal is clearly moving away from zero as if the dark fringe offset is increasing.
An explanation could be that the beam is not well aligned on B1_PD1, and that only ~20% of the light reaches the photodiode.
The shift was dedicated to the continuation of the recovery of the ITF, carried out by A. Allocca. At around 16:00 UTC we started suffering from some instability of the Input Mode Cleaner, A. Chiummo investigated the issue. Also, at 19:30 UTC, the suspensions started to be unstable due to an earthquake ( see attached image ).
The highest step reached stably was LOCKED_OMC1_B1S2_DC. Activity still in progress.
The floor of the sensitivity curve for the gps=1230861618, was generated using the following parameters:
aLF=2.0e-22;
aMF=7.78e-24;
aHF=2.86e-24;
POLE=abs((1i*f+56)/56);
LFN=aLF./(f/20).^4;
MFN=aMF.*(f/100).^0.25;
HFN=aHF*POLE;
THLFN=3.8e-22./(f/10).^2.5;
THMFN=1e-23./(f/40).^.5;
THERM=(THLFN.^2+THMFN.^2).^.5;
DESIGN=(THERM.^2+HFN.^2).^.5;
TOTAL=(LFN.^2+MFN.^2+DESIGN.^2).^.5;
This was used to compare and calculate the gain/loss of Mpcs in BNS range for several peaks.
Today I moved the second Trilium from the "supporting" beam to the "carried" one. Measurements started around 18:30 LC.
The chiller 2 is powered on UPS.
As observed by Irene, we experiment some oscillations of the laser temperature since december. This morning we have reset the PID setup without any improvement so we swaped to the spare chiller (plot 1). Note the PMC power is strongly dependent on the setpoint. It will be decrease asap (plot 2).
The chiller 2 is powered on UPS.
This morning I found the ITF locked in LOW NOISE 1.
The first part of the shift was dedicated to the maintenance activities:
INF Standard Cleaning Operations;
Oxygen deficiency sensors replacement in Terminal Buildings (external firm);
VACUUM Standard Nytrogen Refill;
ENV I.Fiori went in Laser Lab to install and relocated some accelerometers;
TCS Chiller Check performed at 08:45UTC, the refill has been needed only for the NI one;
SUSP Sa_PR cooling unit replacement.
After that, started the recovery and still in progress.
VPM
10:49UTC Group Earthquakes processes restarted after a crash
11:29UTC USGS process restarted after a crash
The SIB2 data sent to the SSFS start to be delivered too late at the SSFS mezzanine level in the same period (see plot)
This is Michal who realigned the OMC, using the single bounce beam from NI aligned.
Yesterday's afternoon was devoted to the recovery of the ITF:
- after a couple of tweaks, both the DIFFp and the PR loop could be engaged automatically during lock acquisition with the usual parameters (gains, setpoints);
- the COMMp loop could be engaged without issues once the other two were closed;
- however this was not possible for the BS loop, I am not sure if this is related just to the working point or there was something different with the signal, to be checked; changing the gains had the only effect to delay or hasten the loss of the control;
- the phases of the RFC and B2 PDs were adjusted.
I found out that the three UPS voltage sniffers located in DAQ room had been erroneously plugged the same multi-plug at the back of the ENV rack. Who did this action probably inteded to "clean up", but indeed the three sniffers must stay plugged multiplugs belonging to different racks in order to correctly monitor the R, S, and T phases of UPS voltage!
So, I plugged them back to the original plugs. Now the signals look correct (Figure 1).
Note that these sensors are also important because are going to be used for 50Hz feed forward.
I did a temporary relocation of two accelerometers and added one:
- relocated ENV_BS_ACC_Z now attached to the turbo pump of the BS-PR link
- relocated ENV_SR_ACC_Z now attached to the tubo pump of the BS-RS link
- added ENV_DAQR_ACC attached to the metallic structure of the outlet duct of the DAQ room air conditioner
Figure 1 shows interesting lines below 50Hz (DAQR) and around 400-500Hz (turbo pumps). We intend to keep them there for a few days to check for possible couplings to ITF.
Note to operators: because of higher rms noise of ENV_BS_ACC and ENV_SR_ACC, some flags could turn on in the DMS: please shelvve them for 1 week.
HRec has been restarted at 9h15 LT to add the subtraction of the CAL_NE,WE_MIR_Z_NOISE channels between h_raw and h channels. The lines sent via these channels as external noise should now be reduced in the hoft_20000Hz channel (37.5, 56.5, 77.5, 106.5, 107.5, 137.5, 206.5 and 406.5 Hz). The lines from PCals are not yet subtracted.
Since the EIB was suspended again, strong 50Hz and harmonic appeared in the strectrum of the accelerometer located on the bench.
Quadrants and photodiodes on the bench are not affected (attached plots).
