On Tuesday (May 17th) was performed a FIS measurement with the squeezing beam going to the retroreflector on SQB1. 

We put the retroreflector into the beam path (SQB1 pico, from 0 to 1.2 million steps) and verified the alignment on EQB1 (PD_MONI, cams & homodynes) by tuning the tilts of the retroreflector, first using the SC (which was preemptively aligned to the FC and in reflection into EQB1 with >3% misalignment, see a previous log entry this week by me (ding)), then using the CC+SQZ beams to maximize the contract on the HD when beating with LO.

Good alignment of CC-LO on HD was reached at 13:23 UTC (check EQB1_HD_DIFF_RF_4MHz for the fringes). 

A first phase scan from 4 to 0 rad was performed at 14:29:00 UTC. Fast gain of CC loop was at 100, coarse at 0.5 (see the corresponding channels in datadisplay). We reached about +6.85 dB of ASQZ and -4.66 dB of SQZ. 

Then the fast gain was increased and a realignment was done to optimize this measurement. At 17:13 UTC a quick scan of the demod phase was done, going from 0 to 3.14 rad in 20s. See plots attached. We reached the following :

- parametric gain g=2.42 (7.7 dB)

- measured antisqueezing A=2.29 (7.2 dB)

- masured squeezing S=0.555 (5.1 dB)

- optical efficiency (g-1)/(A-1)=91%, i.e. 9% optical losses

This nicely compares with the squeezing measurements from EQB1 (at ~6% optical losses), considering the 3% measured extra losses on SQB1 due to clipping/ HWPs in the double pass FIs.

For reference, a shot noise measurement attempt was made around 17:22:20 UTC. Beware to check the OPA_status to make sure that the OPA was locked during the measurement.

Then we started to suffer from OPA unlocks due to humidity changes and we moved away the retroreflector.

NOTE : The Retroreflector either : has a strong hysteresis, or has a problem of getting stuck in the translationnal stage, since we needed to go up to -800000 steps to bring it away (was initially away at 0 step). The mechanism has to be thoroughly checked the next time we open SQB1, since a critical failure on the retroreflector position mechanism means the beams cannot access the FC and FDS is impossible. 

The work of north input CO2 actuator tuning went on in this afternoon shift changing the DAS power and checking the matching scanning the cavities (#55923
), the activity was carried out without any major problem and it concluded at 20:00UTC; I left the cavities locked on the infrared.

DET
DET_MAIN switched state to FAST_SHUTTER_DEAD at around 16:22UTC because of missing data, I restored it back to SHUTTER_CLOSED.

SBE
at 19:15UTC the SQB1 vertical control was opened by the guardian; control closed after re adjust of F0 vertical position with stepping motors.

Here is also the HWS-DET curvature data with the indication of all the steps done today. The figure shows that the best matching is reached when the curvature is closer to zero. The spike before each DAS step is due to the flipper mirror being closed to perform the power tuning.

We leave the NI DAS with the current settings: 1.9 W inner ring; 4.2 W outer ring.

Summary of the work on the North arm, similar to the one on the West arm the day before, see post 55908.

The summary of the day is in the plot attached:

- at the start switch on the CH from 0 to 160 minutes
- then fine tuning of the DAS to restore the matching from 160 minutes to 520 minutes where different setting are tried
- then at 520 minutes switch off the inner DAS to create a big transient to scan fast the CP length (sucessfull to see the change in the HWS)
- based on those data, 630 minutes new setting of the DAS but still not enough recover the matching.

The script to analyse the data have been updated and are in the daily folder

The goal of this simulation is to compute the losses in the FC due to the defects of the mirrors.

The radius of curvature (RoC) of the mirrors influences the losses as some values of the RoC's can create degenerate resonances with TEM00. We want to avoid those RoC's in the FC. Ring heaters are placed around each of the mirrors and can apply a modification of the RoC's between [-10,0] . With this simulation we can check if we need to use the ring heaters to improve the round trip losses (RTL).

The cavity is simulated with Oscar and we used the maps from the TDS (IM and EM) for the mirrors. The current parameters for ADV+ are : 

• FCIM_RoC = 557.9 ± 0.2 
• FCEM_RoC = 559.0± 0.9 
• Average Scattering FCIM = 5.5 ppm () 
• Average Scattering FCEM = 8 ppm ()
• We plot with 50 points betwen [RoC-10; RoC]
• m_diameter = 149.9 ± 0.1mm ;
• Lambda = 1064E-9 m;
• Length_Cavity = 284.9 m;
• Transmission of FCIM : 562 ± 1.0  ppm
• Transmission of FCEM : 3.16 ± 0.01 ppm
• Thickness of Mirror : 89.9 ± 0.1 mm

The Average Scattering is due to spatial defects of frequency higher than the resolution of our maps for the simulation ; thus we need to take them into account. To compute the Round trip losses of the TEM_00 mode in ppm we use : 

The factor 2 takes into account the RTL due to both mirrors.

We make both radii of curvature vary at every step, varying in the range [547.9 ; 559.9]  for FCEM and in  [549.0 ; 561.0] for FCIM (we go 2  above the reachable points for visibility). A more optimal approach currently in development would be a 3D map with X,Y being the RoC's of the mirrors and Z the RTL.

The resulting plot in Fig.1 shows the RTL on the TEM00 mode (orange curve) from the circulating TEM00 in the cavity. The X-axis shows RoC of the IM. To have the RoC of the EM at the same value you need to remove 1.1  . 

At the current settings, the RTL of the 00 mode is estimated at  : RTL = 35.2824 ppm.

This simulation tends to show that we are in a good position for the RoC of the mirrors as we are in the valley of a close peak, although a 3D plot would lead to a better overview.

The shift was dedicated to the planned TCS activity, CO2 North Actuator Tuning, carried out by Nardecchia and Rocchi, with the ISC support of Mantovani.
During the shift, we performed scans while at LOCKED_ARMS_BEATING to verify the Matching of the Cavities.
Activity is still in progress.

DAQ
- SDB2 Vertical Loop opened by the guardian at 2:40 UTC. Vertical position recovered and loop closed at 6:56 UTC. The loop opened again at 9:30 UTC, after further tuning of the motors, the loop was closed at 9:40 UTC.
- SQB1 Vertical Loop opened by the guardian at 2:09 UTC. Vertical position recovered and loop closed at 7:33 UTC.

DET
This morning we found that the camera related to SDB2 (B1, B1p, B5, B1s2) were not available via Telescreen and Gx Client.
I contacted Gouaty via mail about this issue and, thanks to the support of Bonnand, the images were restored at 10:27 UTC.

ITF State: ITF Locked on IR;
ITF Mode: Commissioning;
Quick Summary: IMC and RFC locked; all Suspensions loops closed, SQB1 and SDB2 Vertical Loop open, all SBE loops closed; Temperature in DET_LAB below the threshold; Missing VPM signals from DMS.
Activities ongoing since this morning: No activity known so far in the Control Room.

We have computed the focal length of the DAS lens with the current settings (0.7 W inner and 0.95 W outer) and the result is about 33 km (very close to what estimated in 55895). 

The initial value (0.1 W inner and 2.0 W outer) used yesterday (55898 and 55900) gives a focal length (27 km) off by 20%.

The goal of this shift was to continue the work started yesterday afternoon, i.e. to compensate the north CP cold lens with the DAS power. 

Since last time we checked several values of the outer ring, this time we decided to start from the best point of yesterday (Inner ~70mW, outer ~1W) and to change only the inner power.

To be noted that, as explained after the morning shift, in order to avoid misalignment of PR/BS and arms we perform an unlock/realignment each ~20-30 minutes and we use the new function for the drift control of the arms on the green during all the scans.

A summary of all the several steps performed are shown in figure 1. It seems that the last power chosen for the inner ring (~700mW) improved a lot the north arm matching, which is now close to the west one. Also, it appears that after ~45 minutes we reached the minimum for the north matching . It is maybe worth to fine tune tomorrow morning the inner power around the last chosen value.

In agreement with Alessio, we then decided to leave the arms locked in IR with the North Inner DAS ON.

Below a detailed list of the actions:

We first scan the TEM0 and TEM2 with only the inner NI ring switched on at 14:06 UTC for TEM00 and 14:16 UTC for TEM02

We then put the outer DAS  at 1W: 
1337005078  19 May 2022 14:17:40 UTC  (19 May 2022 16:17:40 CEST)

and we leave the scan of TEM2 going on.

Unlock & cavity realignment
1337006416  19 May 2022 14:39:58 UTC # self unlock
1337009119  19 May 2022 15:25:01 UTC  (19 May 2022 17:25:01 CEST)

Scan of the TEM0 mode
1337009456  19 May 2022 15:30:38 UTC  (19 May 2022 17:30:38 CEST)

Unlock & cavity realignment
1337010062  19 May 2022 15:40:44 UTC  (19 May 2022 17:40:44 CEST)# self unlock

After thermalized, we do a first step on the inner ring, doubling its value from 60 to 120mW:
1337010492  19 May 2022 15:47:54 UTC  (19 May 2022 17:47:54 CEST)

and then we start back the scan on TEM02

Unlock & cavity realignment
1337012028  19 May 2022 16:13:30 UTC  (19 May 2022 18:13:30 CEST)

We do a TEM00 scan
1337013927  19 May 2022 16:45:09 UTC  (19 May 2022 18:45:09 CEST)

The matching seems to improve, thus we decide to double again the inner das up to 250mW:
1337014819  19 May 2022 17:00:01 UTC  (19 May 2022 19:00:01 CEST)

Unlock & cavity realignment
19 May 2022 17:08:30 UTC
1337015565  19 May 2022 17:12:27 UTC  (19 May 2022 19:12:27 CEST)# self unlock

At this time, we experienced a failure while doing the Fmoderr with the automation and we found out that it was due to the hardware. In fact the LNFS (low noise frequency synthesizer, for the 56, 6 and 8 MHz sidebands generation) was in a faulty state. We solved it by switching off and on again but  the data acquired till 17:30 UTC are not reliable.

Once the LNFS problem solved, we start again with a TEM00 scan 
1337018984  19 May 2022 18:09:26 UTC  (19 May 2022 20:09:26 CEST)

We start back to scan the TEM02
19 May 2022 18:17:35 UTC

Unlock & cavity realignment
1337020729  19 May 2022 18:38:31 UTC  (19 May 2022 20:38:31 CEST)

We unlock again to start the new measurement, and we do a TEM00 scan
1337021485  19 May 2022 18:51:07 UTC  (19 May 2022 20:51:07 CEST)

The matching seems to improve, we decide to double again the inner das up to 410mW
1337021506  19 May 2022 18:51:28 UTC  (19 May 2022 20:51:28 CEST)

Unlock & cavity realignment
1337024342  19 May 2022 19:38:44 UTC  (19 May 2022 21:38:44 CEST)

The matching improves very slightly, we decide to increase again the inner das up to ~710mW to see if we can find a value where the matching degrades:
1337027106  19 May 2022 20:24:48 UTC  (19 May 2022 22:24:48 CEST)

We redo the scan on TEM00
1337027210  19 May 2022 20:26:32 UTC  (19 May 2022 22:26:32 CEST)

Unlock & cavity realignment
1337028880  19 May 2022 20:54:22 UTC  (19 May 2022 22:54:22 CEST)

The planned TCS activity on CO2 Actuator tuning keeping the arm matching followup, carried out by Spinicelli, Derossi and Rocchi from remote went on without major problems till 17:00 UTC when a problem with the LNFS occurred.
Chiummo, DeRossi and Nocera (from remote) worked in IER to solve the issue till 18:00 UTC. LNFS box switched OFF / ON. 
Activity resumed at 18:00 UTC and it went on without major problems till the end of the shift.

No Downtime effects on ITF.  Thermalization was in progress. 

Parallel activity:
SGD - Test of the IR-GR resonance using the AOM driven by the DDS (Bawaj, Zhao)

SQZ
21:00 UTC - SBE_SQB1 position loop opened for unknown reasons. Properly closed and restored the vertical position via stepping motors.

Other
(19-05-2022 17:00 - 19-05-2022 19:00) Operator on site with expert from remote
Status: Ended
Description: LNFS failure

Triggered by the current miscentering of the carrier beam on EPRB B4 PC Cam, we noticed that its position changed when the RH was switched on (plot 1).

Figures 1 and 2 compare the image when the arms are locked on IR (as you can see the carrier is now on the top right). To be checked at carm null.

To engage the stray light loop, we need to use the following python command from the SQZ_init.py script:

EQB1_HD_AA.HD_STRAY_LINES_ENBL = 1

and to open it:

EQB1_HD_AA.HD_STRAY_LINES_ENBL = -1

Today we kept the work on the stray light loop. We wanted to close the stray light loop and the CC loops together.

The filter cavity lock precision was ~1Hz. The cavity has been aligned on the IR reaching 14.5 V in transmission.

Stray light loop + CC loops

At 9h43m30 UTC we closed the loops (5 mins of data):

Gain CC fast = 100

Gain CC coarse = 0.5

Gain Stray loop = 5000

CC phase = 0.8 (ASQZ)

In Fig. 1 the signals. We notice that we are introducing a servo bump ~100 Hz. So we tried to reduce the gain of the stray loop.

We took some data, see Fig.2 :

9h50m14 UTC : shot noise (IR shutter closed) in magenta

9h54m04 UTC : CC loops closed (ASQZ, detuning FC 180 Hz), stray light loop open in yellow

9h58m11 UTC : stray loop + CC loops (ASQZ, FC detuned by 180 Hz) in blue - we notice that we are killing the FDS hole using the stray light loop.

We decreased the gain of the stray loop from 5000 to 4500 (10h07m12 UTC). There was an interplay between the two loops, introduced by the extra phase noise introduced by the stray loop dither line at 500Hz.

500Hz line compensation with M1 LO

So we decided to go on with the compensation of the 500Hz line on M4, acting on LO_M1. We had lots of troubles due to the micro seismic activity, that was introducing a phase shift too large to be corrected with the range of our actuators (CC fast on AEI shifter and CC coarse on M1 and M2).

We set the amplitude of the dither line on M4 at 10mV, to avoid sending a line too large on M1. In fact with 15mV we needed to set ~200mV to L1 to compensate the line, and it was unlocking the CC loop.

The final parameters are:

• Dither line M4_Z @ 500 Hz, 10mV
• Compensation line LO_M1 @ 500Hz, 80mV.

The compensation of the line is shown in Fig.3. We are limited by the line at 2kHz, whose origin is not clear to us. It is also present in the MAIN PLL channel. We will further investigate it in the next days.

Tune of the stray light loop gain to avoid the servo bump

We further decreased the gain of the stray light loop and we took some data:

• Gain = 1500 @ 14h20m57 UTC, detuning at 180 Hz, stray light loop closed and CC loops closed for 5 mins -  in Fig. 4 and 5 we see that we are touching the shot noise (yellow it is only CC loops on)
• Gain = 1200 @ 14h25m20 UTC, detuning at 180 Hz, stray light loop closed and CC loops closed for 3 mins

We kept the gain = 1200.

The line at 500 Hz on the CC error signal channel, was oscillating because the correction of M1 was moving a lot. So we turned off this actuator, using the M1 only for the compensation. It was better.

Stray loop + CC loops at different detunings

• detuning 130Hz @ 15h24m00
• detuning 80Hz @ 15h24m47
• detuning 160Hz @ 15h27m20 (1 min)
• detuning 100Hz Fig. 7
  • stray ON + CC ON @ 15h28m44 (magenta)
  • stray ON + shot noise (only LO) @ 15h37m04 (2mins, yellow)
  • stray OFF + shot noise only LO) @ 15h39m24 (2mins, green)
• detuning 50Hz Fig. 8
  • stray OFF + shot noise (only LO) @ 15h43m30 (magenta)
  • stray ON + shot noise (only LO) @ 15h45m00 (yellow)
  • stray ON + CC ON @ 15h48m50 (green)

We conclude that

• the loops are working together;
• we need to tune the filter of the stray loop in order to see the hole on ASQZ when it is closed: because we are introducing a small servo bump at tens of Hz;
• we are no longer exciting peaks at tens of Hz when closing the stray light loop.

In the continuity of our last entry (#55824) we investigated the influence of the PR centering activity on the relative positions of the carrier and the sidebands. To perform this analysis we used the B4 phase camera (V1:EPRB_B4_PC_56MHz_USB_PosX/Y and the equivalent channels for carrier, USB/LSB sidebands at 6MHz, 56MHz). We compared the situation before the PR centering intervention (#55627) and after. Since the 0 of the camera is arbitrary and changes sometimes during resets or commissioning operations, we considered the PosX/Y with respect to the carrier (position 0 corresponds to the carrier beam center).

FIg 1 & Fig 2 shows the distribution of the difference between carrier and sideband PosX/Y before and after PR change (20/03/2022 – 25/04/2022 & 29/04/2022 – 17/05/2022). For each lock we took the average value between 500 & 1300 s (permanent regime is reached).

We see that the distance seems to have changed by the same amount on X and Y, whereas in (#55824), dividing the PRCL amplitude by the PR tilt dither amplitude seemed to indicate that the motion on X-centering had been about one order of magnitude smaller than the one on the Y-centering.

Fig3  & Fig 4 shows the distribution of the radial distance between the sidebands and the carrier.

We tried using a longer time interval for the pre-intervention situation in order to have a larger number of locks (and more statistically meaningful results); however since January 2022, a number of activities seem to have largely affected the respective positions of the carrier and sidebands. Of course, the number of locks after the change  is much lower than before, so that the statistics are not extremely reliable.

In addition to the online overlap channels, we computed the distances (sqrt(x²+y²)) between carrier and sidebands / between both sidebands of the same frequency.

In conclusion, according to the histograms :

• On X, the sidebands used to be well superposed with the carrier. They are now still overlapping between them, but moved further away from the carrier.

• On Y, the superposition between the sidebands is now better and the shift with respect to the carrier has been greatly reduced. However, they still do not match perfectly either between themselves or with the carrier.

• On the whole, the distribution width of sidebands and carrier on the mirror is almost the same (or slightly reduced), but now distributed evenly between X and Y (see absolute distance histograms).

(The x axis has the same scale on each of the histograms.)

The results on our dithering analysis are not exactly comparable between the X- and Y-directions of mis-centering, since we normalise by the MAR amplitude and not by the MIR one. The proportionality between marionette and mirror amplitude is slightly different on TX and TY. Therefore, we have also extracted the mirror dithering amplitude. Dividing PRCL (Z length) amplitude by mirror (angular) amplitude gives, up to a gain which is now the same on X and Y, an estimate of the actual coordinates of the beam center on the mirror.
The mean results are :

• [fig 1] before intervention : X 0.07 Y 1.1
• [fig 2] after intervention : X 0.1, Y 0.3

They are to be taken with caution, the error margins being quite wide. They give however a comparison between X and Y on the same scale.

nb : left column of figures is entitled 'TY', but actually corresponds to 'Y-centering'.

mistake: DAS and CH off (not RH)

This morning we have worked on the TCS tuning for the Warm.

Starting configuration: DAS and CH off (initial data points in Figure 1) mode matching value 0.0148.

@ 8:10UTC the WI CH power has been set to 93mW to mimic the YAG beam (~0.572W on the pick-off powermeter), which corresponds to the beginning of the first yellow circles in Figure 1.

@ 9:46UTC the WI DAS has been turned on (0.4W IN 1.1W OUT, values expected from simulations). A bit before than the light cyan circles, see Figure 1.

The Figure 1 shows the matching evolution starting from 7:38:30 UTC. The 00 mode has been measured @ 7:54:40UTC and @ 14:06:15UTC giving respectively : 23681.7038,  23637.409 (values to use for the 02Mode normalization).

It is visible that after the DAS compensation the matching is back to the initial value, which means that the 0 lens is achieved.

It is visible that the matching is well in agreement with the circulating power in the cavity, see Figures 2 and 3. (the last data points are not compared because the agreement was not so optima, which is most probably due to etalon effect).

Side note:

- the north cavity was badly misaligning during the scan (with drift control off). It has been improved by engaging the Drift control on the cavity mirror. Same situation for some data set of West cavity.

The command to act on the drift control (no PR translation and BS tilt) is drift_control_TMs(1) to engage and drift_control_TMs(0) to disengage.

This morning shift was dedicate to tune the west input CO2 actuators, before starting with the activity we took a reference with this two scans:

scans on TEM00: GPS 1336982046
scans on TEM2: GPS 1336982306

the work was carried out by Ilaria and Maddalena changing the CO2 actuator configuration while scanning the arms, the activity went on all the shift without any major problem with the exception of some difficult in the locking on the north green laser; during the shift we experience the effects of this earthquake:

M 6.9 - Macquarie Island region
Time 2022-05-19 12:13:31 (UTC+02:00)
Location 54.137°S 159.084°E
Depth 10.0 km

this didn't impact on the activity of the ITF but it spoiled the SQZ works, the TCS CO2 tuning will go on in the afternoon...

other activities:

- SQZ: activities on the SQZ system.
- ENV: newtonia noise sensor installation and commissioning.

DMS
Since this night at around 3:00UTC the VPM DMS section was no more updated (grey flags on DMS), under investigation.

Yesterday afternoon we tried to work on the stray light loop in order to make it work with the CC loops on.

We managed to close the stray light loop only with the LO, acting on M4 (see Fig. 1).

Unfortunately due to humidity problems in the lab, the OPA was not staying locked so we couldn't go on with the work.

The goal of the shift was to tune the DAS at the NI in order to compensate the CP converging lens on the North arm.

We started from the configuration prepared during the morning shift, that is the matching of the West Arm with the TCS off.

Then, Claudia switched on DAS on NI with the following parameter:
15:25 UTC: NI_DAS on with 0.1 W inner and 2 W outer

The thermalization is supposed to be longer than 2h, we then left the scan of TEM2 going on for this time. At the end of this period, we found a huge misalignment that probably spoiled the last  measurements. 
The analysis of the first DAS point (fig. 1) shows that we have an improvement in the first part followed by a degradation of the matching. Anyway, the north arm couldn't reach the same level of matching as the west one. 

We then decided to take a scan of the FSR/TEM2 before and after a realignment of the arms in order to verify if there was a difference in the estimated mismatch. 
In fig.2 we see that the realignment  improved the matching in the last points.

From now on, we block the scan and realign each ~20 minutes to avoid this problem. In agreement with Alessio, we decided to look at another DAS point by changing only one actuator. We thus increased only the outer ring power:

1336931077  18 May 2022 17:44:19 UTC  (18 May 2022 19:44:19 CEST):  doubling outer das (from ~2W to ~4W)

During the thermal transient, we unlocked and realigne few times:
1336932459  18 May 2022 18:07:21 UTC  (18 May 2022 20:07:21 CEST)
1336934045  18 May 2022 18:33:47 UTC    (18 May 2022 20:33:47 CEST)
1336935289  18 May 2022 18:54:31 UTC  (18 May 2022 20:54:31 CEST)

As the matching seems to go into the wrong direction (fig. 3), we decided to reduce the power instead.

19:02 UTC: outer ring from 4 W to 0.95 W

During the thermal transient, we unlocked and realigne few times:
1336936797  18 May 2022 19:19:39 UTC  (18 May 2022 21:19:39 CEST)
1336937565  18 May 2022 19:32:27 UTC  (18 May 2022 21:32:27 CEST)      (unwanted => green unlock?)
1336939657  18 May 2022 20:07:19 UTC  (18 May 2022 22:07:19 CEST)

The matching indeed improved (fig.4).

We continued switching off completely the outer ring:
1336940042  18 May 2022 20:13:44 UTC  (18 May 2022 22:13:44 CEST)

During the thermal transient, we unlocked and realigne few times:
1336941248  18 May 2022 20:33:50 UTC  (18 May 2022 22:33:50 CEST)
1336942679  18 May 2022 20:57:41 UTC  (18 May 2022 22:57:41 CEST)

Apparently, this step didn't improve the matching (fig5, fig 6).

A summary of the different steps is shown in figure 7.

The DAS tuning will continue tomorrow.

We leave the arm locked on IR and the only WI inner DAS ON @0.1W.

BRMSMon recompiled against Fd v8r43p3, as suggested by Benoît. New tag created (v6r0), VPM configuration updated and server restarted.

The planned activity of TCS proceeded for all the shift without issues; the activity consisted in locking the ITF at LOCK_ARM_BEATING and the manually launching the scan. Activity still in progress...

The problem at the humidity of DET prevented the SQZ activity.

ACS

no info about the intervention of the exneral firm started in the morning.

After the work on the north arm (#55889 and #55886) this morning we made the matching for the west arm with the TCS off (DAS and RH).

We did the following steps:

• 1336903483 10:04:25 UTC scan of the arms with the green lasers CARM_SET_scan(3)
• 1336904178 10:16:00 UTC L1z 20000 Backward
• 1336904358 10:19:00 UTC CW 10000 Backward
• 1336905198 10:33:00 UTC scan of the arms with the green lasers CARM_SET_scan(3)
• 1336905918 10:45:00 UTC L1z 20000 Backward
• 1336906158 10:49:00 UTC CW 20000 Backward
• 1336905198 10:58:00 UTC scan of the arms with the green lasers CARM_SET_scan(3)
• 1336907418 11:10:00 UTC L1z 20000 Backward
• 1336907658 11:14:00 UTC CW 20000 Backward
• 1336905198 11:22:00 UTC scan of the arms with the green lasers CARM_SET_scan(3)
• 1336909103 11:38:05 UTC scan CARM_SET_ONmode(0,2)
• 1336909222 11:40:04 UTC scan CARM_SET_scan_12(2)
• 1336909809 11:49:51 UTC L1z 20000 Forward
• 1336910645 12:03:47 UTC scan CARM_SET_scan(1)

In conclusion, we moved the L1z by 40000 steps backward. Analysis will follow.