Everything has been temporarily stored in the Laser lab waiting for the completion of the works in the Optics lab.
Next week, we should move everything in the optics lab.
I have attached a few pictures of the EIB showing the bench before and after the optics have been unmounted.
The 4 fibers of the laser amplifier have been put below the laser bench, the 2 fibers of the slave laser has been disconnected from the slave laser head and left on the bench (they should be safe from the point of point of the activity on the acoustic enclosure since they are protected by a ringed electric sheath). The phase camera reference beam fiber has been removed from the bench and stored in the half rack close to the laser bench.
Tomorrow before starting the activity of unmounting, vacuum people should vent the Pre-Mode cleaner cavity.
In order to realign the beam towards the detection we moved the beam splitter. We found out that the beam was mainly tilted in the vertical direction by more than 500urad. In order to recover this tilt we moved the SIB in tz(around the IMC axis) and recovered the alignment on the IMC by adjusting by hand the beam pointing on the EIB.
Once done we have to realign the beam on IMC AA quadrants in order to be able to close the AA.
The next step consisted in looking at possible flashes in the north arm. We played around with the beam pointing (using SIB_M6) and realigning NI and NE mirrors.
Again the beam had to be tilted by about 200urad vertically. In this configuration we could get flashes of intensity on north end bench of about half of the usual power transmitted by the end mirror at step 1. Note that the final adjustment of the beam pointing will be done tomorrow.
At that point we worked on the realignment of RFC cavity using M13 and M14 mirrors. Flashes of intensity of about half of the usual power have been obtained. The final alignment will be made tomorrow.
To complete the activity, we have still to:
- recover 200urad of vertical tilt of beam direction.
- relock and fine tuning of RFC alignment.
- move BMS quadrants on the new reference position of the beam respect to the ITF.
First we recentered the EIB Faraday isolator and the IMC mode matching telescope optics.
Then we realigned the IMC cavity and could easily lock it as soon as the 22MHz quartz oscillator was put back in place.
IMC automatic alignment quadrants have been been recentered (note that there is a problem with NFv translation stage that we were not able to move. To be investigated...).
Once we were happy with the IMC cavity alignment, the BMS quadrants have been realigned on our new reference position and we could easily close the BMS loops.
During the work we have discovered that RFC tra photodiode has been for some unknown reason unscrewed form its bench. Tomorrow we should align properly the RFC setup to relock it.
First we have readjusted a bit the matching on PMC cavity so that we have about 33 W transmitted by the PMC.
Then we realigned the beam in the EIB Faraday isolator and in the IMC Mode matching telescope by using LB_M13 and LB_M14 mirrors of the laser bench.
The consequence is that we had to shift the 2 EOMs by a bit less than a mirror to recenter them on the main beam.
Then we started to realign roughly the beam on BMS quadrants and start to play with SIB and MC positions to realign the IMC respect to the laser beam.
The alignment of the IMC is quite good. Fashes of power up to 10W are visible on IB_TRAMC photodiode.
The next step is to retune the IMC loop to relock the cavity.
We left the system up and running for the week end. The PMC cavity has been left in scanned mode so that the main beam is dumped on the water cooled beam dump of the bench.
Here are the last fonctionning parameters:
Amplifier:
diode temperatures 21,21,25.7,24
pumping currents: 52,45.7A /chiller 18
SL
pumping settings:
PS3 set 13.1-29A
PS1 set 13.1-29A
The CHRoCC was tested under atmospheric pressure and the heater was set to 440 C (2.5 V, 5A): the surface of the window was heated until the heating pattern given by the IR was stable (figure 2). The amplitude of the temperature of the heating pattern is about 2 C.
Then a circle of the dimensions of the IMC mirror was printed thanks to the heat of the fingers. It makes the reference for the data analysis (figure 3).
Profile of the experimental heat pattern is compared to the expected results from Zemax simulations (figure 4). The shapes of the temperature profile and the intensity profile are matching, that is a first confirmation of the good behavior of the system. The system is now ready for an installation in IMC tower.
This work has been completed at the beginning of the afternoon and now the laser and the injection system are back in operation.
We are ready for an installation in IMC tower from next Tuesday on.
On the detection bench we made a fine alignment of the telescopes on both the paths (B1p and B5) in order to obtain on the PC photodiodes the image planes of the NI mirror. This was done by introducing a slow and continuous angular movement on the NI mirror and we tuned the distance between the lenses of the telescopes in order to minimize the movement on the cameras.
In the figure the names of the channels are always swapped.
Figure 1 shows the images (amplitude and phase) of B5 carrier and sidebands.
Figure 2 shows the images (amplitude and phase) of B1p carrier and sidebands (LSB image has electronics troubles).
We decided to leave PC B5 connected to B1p electronincs in order to acquire both the sidebands in the next measurements. Therefore we swapped the names in the configuration of the ADC.
We cleaned a few mirrors and lenses on the beam path, realign B2 photodiodes and dump properly spurious reflection coming from the telescope lenses (EIB_B2_L1 and EIB_B2_L2).
Then we check the centering of EIB faraday isolator since the matching on the IMC cavity was quite bad. The beam looked a bit low so we realigned the beam in the Faraday. The consequence was that the matching improved a lot we went from 23% down to 12.8% (see attached plot).
After relocking the ITF we could see that our intervention got a good effect on the sensitivity the structure around 230Hz disappeared.
Once done Paolo realigned properly the input beam and the ITF. Locking activity could start at the beginning of the Afternoon.
We reached quite easily step 12 and could strat to compare the sensitivity with the new support repesct to the old support. NE Chrroc has been changed from 3.54V to 3.58V.
On the first plot, one can see the comparison of SSFS corr signal with old support (purple) and with the new support (black). As expected, mainly 3 peaks are removed (55, 78 and 82 Hz). New structures appeared starting from 160 Hz on. It is not sure taht everything is coming from the new support. It can also be that some resoannces coming from the bench for example are appearing now because there is no filtering anymore.
We did a few experiment to check the improvement.
On the second plot, you can see the comparison of the sensitivity curve on Sep 19 (black curve) respect to the sensitivity we had we a mistuning of the 6MHz modulation frequency of about 5Hz (red curve is with new dihedron support and blue with old dihedron support).
Plot 3 and 4 show some zoomed view around 70 and 240Hz.
On plot 3 we can see that 55, 78 and 82 Hz peaks disappeared and on plot 4 we see a structure around 240Hz taht appeared.
The structure around 230Hz is proabbly due to some mount resonances on B2 optical setup (tomorrow we will properly realign B2 setup).
Fig. 1 shows the response of the real (Suprasil) dihedron on the old support, Fig. 2 on the new support, for acc. placed in the centre of the dihedron, bottom edge (green) and upper edge (blue), with respect to the bottom plate (red). The resonance at 8 Hz which appears in all spectra, is due to the table and is of no further relevance here. On the old support, one sees strong resonances at about 50 and 70 Hz. On the new support, this frequency region is smooth; there is, though, a structure starting at ~100 Hz and rising until ~150 or even 180 Hz with a Q of ~ 4 .
The following measurements have been taken with the aluminium (?fake?) dihedron, which has a slightly higher mass than the suprasil dihedron. Fig. 3 shows the response of the fake dihedron on the new support, and Fig. 4 shows the same when the dihedron is loaded with an extra 2.16 kg of mass. This extra mass leads to lower resonance frequencies, which go from ~ 150 Hz down to 60 and 80 Hz. Whereas this qualitatively confirms that this structure at 150 Hz is due to a spring effect, it seems to be with roughly a factor of two too much. We speculate that this is the spring effect of the contact points due to the elastic deformation, which the aluminium suffers at the contact on the steel spheres. However, as this spring constant changes also with the deformation, we are not sure whether our speculation holds. We note that with the old support, we do not see specific sharp resonances at 78 and 82 Hz. It is possible that these modes are not visible by the accelerometers in the longitudinal direction. These points have to be investigated further.
In order to relock the IMC cavity we had mainly to tilt the SIB by 250urad.
This change in tz had the consequence to misalign IMC far field quadrant.
We realigned this quadrant by moving a mirror EIB_ReMC_M4 since the translation stage used to move the quadrant was at the end of its range.
In order to compensate this tilt that induced a misalignment of the RFC we first moved SIB_M1 a mirror just after the dihedron and in order to recover the alignment respect to the ITF we moved SIB_M6 mirror by about 150 steps.
NB:
SIB_M1:
old position:
tx=-1430 ty=-8630
new position:
tx=-2910 ty=-8600
SIB_M6:
old position:
tx=-1707 ty=-2932 tz=-1305
new position:
tx=-1562 ty=-2902 tz=-1305
On the first picture you can see the dihedron on the new support.
Then we performed a fine balancing of the bench. The balancing was difficult since the balancing was changing by itself when we were moving the bench changing the position of small masses on the bench. We finally succeeded to have a reasonable balancing and we decided to close the tower.
Concerning the possible error on ty angle at the level of the dihedron, computations have been done to understand what is the situation after the dihedron support change.
Attached plot gives the difference of the signals that we monitored as reference for the dihedron position and tilt (IMC_QF_DCh,IMC_QN_DCh,IMC_ref_PosX) for the same alignment of the bench before (purple) and after (black) the support substitution.
The difference seen on IMC_QF_DCh, IMC_QN_DCh, IMC_ref_PosX are consistent with a tilt (ty) of the dihedron no more than 100urad.
Phase 1: Mechanical transfer function measurements with new and old support and dihedron/fake dihedron have been performed.
Those results will be reported in a dedicated logentry that will be posted in the next days (see picture 1: the dihedron on its new support during the TF measurement).
Phase 2: Once transfer function measurements were considered as satisfactory, we decided to reinstall the dihedron on the SIB.
On picture 2 and 3 you can see the dihedron on its new support during the reinstallation phase. During the installation we encountered a small problem in the sens that the new support was a bit larger than the old one (probably 100 or 200 um). The consequence was that we had to remove our lateral local reference (installed yesterday) because we could not recovered the shift of the dihedron induced by this small difference. The main effect is that the dihedron is shifted by a few tens/hundred of microns along the input beam path. This small shift is considered not to be a problem.
Then we removed the weights and freed the bench. At that point, we performed a first balancing of the bench and Paolo could close properly the IB LC system.
Then we send the laser beam in INJ tower and we gave a look to our reference signals (IMC_ref quadrants and camera). Everything looked fine except that we can not disentangle dihedron x and ty positions easily from our signals. Tomorrow morning we will do a bit more work to understand if everything is fine.
The last activity we will have to perform tomorrow morning consists in adding 3 more screws on the dihedron support to be sure that the central part of the new support is well in contact with the bench plane. Then we will perform a final balancing before giving the green light to vacuum team for INJ tower evacuation.
After this activity, Irene Fiori and Thomas Bauer arrived and started to install the mechanical transfer function measurement setup in order to have everything ready for the measurement we will do tomorrow morning.
NB: we measured the weight of the old and new support,
Old support: 862g
New support: 821g
then we shorted the telescope distance by 4mm (now LB_L6-LB_L7 distance is 10,3 cm ). We finished by optimizing amplifier pumping config.
Finally we obtained a modest improvement (see attachment). Note that the operation also improved IMC matching as side effect.
We moved IMC_Ref camera by about 25cm towards west direction as expected. The optical setup on the beam path has been slightly modified since we had some problem of interaction between B2 quadrant and this camera.
On the second figure attached, you can see how the beam looks like now of IMC_Ref camera.
Everything will be put back once the new support will be installed and the injection system successfully relocked.
All baffles are done in 6H-SiC manufactured by Boostec; surface quality is "rectified" (no optical polishing). The baffles were AR coated on one side by CVI. AR quality was measured with an integrating sphere at the 2% level. Transmission is zero.
Here is the geometry of the different baffles (see drawing)
"Baffle A": internal diameter 70mm ,10mm thick
"Baffle B": internal diameter 10mm ,4mm thick
"Baffle C": internal diameter 3mm ,4mm thick
