NAOJ GW Elog Logbook 3.2
[Takahashi, Aritomi, Yuhang]
Since BS suspension motion was too large and the pitch picomotor can move in only one direction, we opened BS chamber and checked suspension. Takahashi-san found that the magnet was touching the coil holder so he adjusted it.
Regarding the picomotor, the picomotor cannot move in one direction because the required force to move the picomotor in the direction was too large. To reduce the required power to move the picomotor, Takahashi-san put a spring (Fig. 1). Then the picomotor can be moved in both directions.
We aligned PR and BS to make the green beam at center of the GV between input/BS and the first target.
Before closing the chamber, we checked BS oplev transfer functions (Fig. 2,3) and response of each coil. The transfer function and coil response are fine.
Finally we closed the chamber and started the evacuation of BS chamber.
Dan Chen, Marc
Today we assembled on the 75 * 90 cm optical table the frame of the dark box inside which we will perform the scattering measurement.
We also placed AZTEC #1 and #3 on holders inside the frames.
Report about AZTEC #4 absorption.
It seems to be around 100 ppm/cm and some fishbone patterns are visible as well.
A proper beam dump thorlabs LB2 (1um - 12um) is installed for the newly replaced BS mount (as shown in Fig.1).
Although we can see some green light, let's remind that it is because the camera sensor is more sensitive to green as our eye. There are proper Dichroic mirrors to remove them.
Yuhang, Aritomi, Marc
A new PD (PDA36A-EC) is used (gain is chosen as 0dB), firstly DARK NOISE is measured (saved as DARKL.txt and DARKH.txt)
A LA1608-YAG lens (75mm) was always used when the beam is apparently large.
1. Measure Mephisto RIN. It is measured at the location where we pick off main laser for CC PLL LO.
The DC voltage is measured as 2.66, 2.61V and 2.67V (measured before and after spectra measurements). (saved as MAINL.txt and MAINH.txt)
2. Measure homodyne LO RIN. It is measured just after IRMC transmission.
The DC voltage is measured as 800mV, 780mV (measured before and after spectra measurements). (saved as LOL.txt and LOH.txt)
3. Measure SHG generated GR RIN. It is measured before MZ.
The DC voltage is measured as 3.14V, 3.16V, 3.24V (measured before and after spectra measurements). (saved as SHGL.txt and SHGH.txt)
This is an experiment record. An analysis of data will come later.
First we replaced the broken BS oplev PSD and measured BS oplev spectra as shown in the attached figure. The noise floor of BS oplev spectra is fine, but BS suspension moves a lot. In addition, BS pitch can go down with picomotor, but cannot go up. We need to open BS chamber again...
We somehow aligned FC and could see the green flash, but green beam spot at FC transmission moves too much due to BS suspension motion.
Then we found that we cannot maximize TEM00 very well and yaw misalignment cannot be removed.
We aligned green injection in yaw before as reported in elog2940. We doubt that this could cause some problems. So we tweaked the green injection, but the green flash still had yaw misalignment. After we brought the green injection back to the original alignment, END yaw was misaglined a lot and we had to move the END yaw with picomotor. We went to END room and moved END picomotor with a joystick. However, at some point, END yaw picomotor stopped moving and we lost green flash...
To do list for END mirror:
- center END oplev and check END oplev spectra/transfer function
- install a new picomotor driver to avoid going to END room for picomotor control
- If END oplev spectra are strange or picomotor does not move with the new picomotor driver, open END chamber and check the suspension and picomotor
The OPO pump relative intensity noise was measured at different power level.
1. MZ offset 4.2 (25 mW), DC voltage on oscilloscope 320mV
2. MZ offset 4.4 (35 mW), DC voltage on oscilloscope 441mV
3. MZ offset 4.0 (15 mW), DC voltage on oscilloscope 186mV
As a reference, in LIGO, the SHG RIN is about 3e-6 from 3Hz to 400Hz. Then it gradually increases to 3e-5 at 4kHz. This indicates that the control servo has 1/f slope. (alog45088)
So we have a better RIN at low frequency (limited by PD dark noise). But a worse RIN at kHz region.
Yuhang, Aritomi, Marc
A new PD (PDA36A-EC) is used (gain is chosen as 0dB), firstly DARK NOISE is measured (saved as DARKL.txt and DARKH.txt)
A LA1608-YAG lens (75mm) was always used when the beam is apparently large.
1. Measure Mephisto RIN. It is measured at the location where we pick off main laser for CC PLL LO.
The DC voltage is measured as 2.66, 2.61V and 2.67V (measured before and after spectra measurements). (saved as MAINL.txt and MAINH.txt)
2. Measure homodyne LO RIN. It is measured just after IRMC transmission.
The DC voltage is measured as 800mV, 780mV (measured before and after spectra measurements). (saved as LOL.txt and LOH.txt)
3. Measure SHG generated GR RIN. It is measured before MZ.
The DC voltage is measured as 3.14V, 3.16V, 3.24V (measured before and after spectra measurements). (saved as SHGL.txt and SHGH.txt)
This is an experiment record. An analysis of data will come later.
A comparison of RIN, shot noise and photo detector dark noise is shown in the attached first figure for several locations on the squeezer bench. Other figures show these noises for different single location separately.
With help of Yuhang we removed #3 from the translation stage.
Then, I measured the bulk reference sample at the usual position and got R = 0.611 cm/W.
I installed the 1.5 inch sample and measured its absorption at the longitudinal center.
I measured about 35 ppm/cm which is compatible with old measurements.
Finally, again with Yuhang help, we installed the #4 and started absorption measurement.
Marc, Yuhang
Yesterday I tried to remove the 4 screws holding one of the 2 crystal.
Once they are removed, it is possible to access 2 set screws that fix the linear translation stage.
The 4 screws were really tight and I only manage to remove 3.
With Yuhang help, we tried to remove the fourth one without success.
In the end, we broke a screwdriver head inside the screw..
Fukushima-san from ATC helped us to remove it today.
After that the linear translation stage was replaced with a motorized one and the Soleil-Babinet is now closed with 3 out of 4 screws (need to purchase the broken one).
Yuhang and Michael
We replaced the mirror mount for the beam splitter that goes into the SHG. The new mirror mount is much smaller, allowing us to more easily remove ghost beams and scattered light from the squeezer bench.
The beam splitter was replaced by roughly using an iris, SHG opening and GRMC EOM as beam position references. First, the beam splitter was moved by hand to align to the references. Fine tuning allowed us to recover the SHG mode matching to slightly better than the previous alignment condition.
Reference SHG alignment is shown in figure 1 and 2. BS model number is in figure 3. The mirror mount type is shown in figure 4.
A proper beam dump thorlabs LB2 (1um - 12um) is installed for the newly replaced BS mount (as shown in Fig.1).
Although we can see some green light, let's remind that it is because the camera sensor is more sensitive to green as our eye. There are proper Dichroic mirrors to remove them.
Marc, Yuhang
Today the AZTEC #4 was delivered.
We brought it to the PCI cleanroom and applied first contact on one surface.
[Takahashi, Aritomi, Marc, Yuhang, Michael]
Since the BS picomotors didn't work, we opened BS chamber and checked the picomotor. We found that BS was stuck at the stopper because we moved BS picomotor too much. After Takahashi-san adjusted suspension, BS picomotors worked well. Before we close the chamber, we aligned green. We moved PR to make the green beam at the center of BS and moved BS to make the green beam at the center of first target.
We also found that the BS oplev beam is hitting off center of BS. We tried to make it at the center of BS, but the reflection was hitting the pillar of BS suspension so we left it as it is.
Finally, we closed the chamber and started the evacuation.
Some notes for future alignment:
- We should not move picomotor ramdomly. We should move the picomotor little by little by looking at some references. We should not lose the green beam at the first target as much as possible.
- When we have the suspension or picomotor issue and open the chamber, we should make sure the green alignment at GV between BS/input and first target before closing the chamber.
- We should measure the oplev spectra/transfer function and each coil response before closing the chamber
[Aritomi, Marc]
At the beginning of today, the BS picomotor was working well. However, after some BS alignment, BS picomotor does not work again... We don't know the reason but maybe because we moved the BS picomotor too much. We need to open BS chamber again.
There were some mistakes in the analysis so these results are not correct.
I'm also starting to doubt the way we analyze our measurements so I'm thinking to speed up the installation of the Soleil-Babinet compensator for this characterization.
[Takahashi, Aritomi, Marc]
We replaced the two broken BS picomotors. After the replacement, we confirmed that the picomotors are working well. We started the evacuation of BS chamber. We will open GV between input/BS tomorrow.
This is a report of AZTEC #3 birefringence measurements where each map had for parameters : X_center = 400.735 mm, Y_center = 122.105 mm, Z_center = 70.275 mm, with 70 mm radius.
Birefringence characterization :
Figures 1 to 6 are the usual polarization measurements, 7 to 9 summarize informations about the stress tensors and delta n and theta distributions.
The sample seems quite uniform and is now a good candidate for KAGRA (it also had quite low absorption).
c axis orientation estimation :
In figure 10, I tried to compute from the mean output polarization the orientation of the sample c axis.
The mean output polarization is fitted by a sine square as a function of the input polarization where errorbars represent 1 sigma uncertainty.
From this fit, it seems that if we inject 50.23 deg input polarization we should have the same at the ouput.
In order to constrain a bit more the fit I am now doing a measurement with p polarization at the input and then will start the measurement at the polarization angle that should be conserved throughout the sample.
There were some mistakes in the analysis so these results are not correct.
I'm also starting to doubt the way we analyze our measurements so I'm thinking to speed up the installation of the Soleil-Babinet compensator for this characterization.
Marc, Matteo
We installed the AZTEC #3 on the translation stage.
We tuned the input laser power to have better SNR on the PSDs.
With about 2 mW power and 3 dB attenuator before the lockin amplifier we are roughly at half the PSDs saturation and about 0.6 V in the lockin-amplifier (saturation at 1 V).
Measurements on-going (one map is 5h40 and we need at least 6)
Following work in 2958, I had to tune the position of the input oplev tilt PSD to have enough space to move the shift PSD on its rail.
To perform decoupling between shift and tilt, I injected a line at 0.94 Hz (length mechanical resonance frequency) with 200 counts amplitude on input mirror length.
I moved the lens on the rail, for each position recenter the beam on the tilt PSD and measured the peak amplitude at 0.94 Hz.
Result is reported in fig 1.
It seems that for pitch, the ideal lens position should be about 12.2 cm while it should be about 11 cm for yaw.
For yaw, the peak was always present and as suggested by Matteo could be caused by the easy excitation of yaw when exciting length.
EDIT : corrected z units from mm to cm.
We put HWP with rotation mount before compensator and try to measure retardation of HWP as entry 2965.
We maximize p-pol power to rotate the HWP:
s = 19.9 mV
p = 648 mV
We rotate HWP and compensator to 45deg.(same direction)
we maximize p-pol power again and record value of the Vernier screw.
d = 6.84 mm
s = 766 mV
p = 16.8 mV
P-pol power was too low. It should be pure p-pol but p-pol power did not change even though we rotate the Vernier screw.
I think polarization direction was parallel to fast(slow) axis of compensator.
Retardation of HWP N_hwp is given by
N_hwp = 1-6.84/13.365
= 0.488
13.365 is calibration parameters measured in entry 2965.