We performed the PF oplev centering.
When centering the QPD of the PF TILT OpLev, the beam did not enter the center of the QPD even though I moved it to the limit of what I could move with the micrometer. Therefore, the fixation between the QPD and the optical surface plate was loosened, and the QPD was moved for centering.
Although centering was performed this time, it is possible that the beam position may shift as the suspension system cools down.
Therefore, it will be necessary to monitor the position of the beam spot in the future, and if necessary, re-centering after the cooling of the suspension system is completed.
We performed the PF oplev centering.
During the data collection for Oplev, the value of the SUM count for IY PF Oplev was found to be very low, about 100.
Since the S/N ratio was poor as it was, the cause was investigated.
The power of the laser on the QPD side was measured with a power meter, and the power on the QPD side was very low at about 2µW.
I looked inside from the view port and could not see any major scattering.
I then measured the power on the injection side with a power meter. The laser displayed 1 mW, whereas the value measured with the power meter was about 6 µW.
Even when the laser power was increased to its maximum value of 2.89 mW, the value measured by the power meter was about 25 µW.
So the laser was replaced with a spare unit, and the TILT QPD was about 11,000 counts and the LENGTH QPD was about 7,500 counts.
Since this value is comparable to other Oplevs, we conclude that the low value of SUM for PF Oplev is due to the weaker output of the laser of the light source.
Based on Mr. Ushiba's advice, I also investigated the effects of turning the power supply on and off and the effects of optical fiber by using a different optical fiber on the optical surface plate near IX, and the results were not different from those measured at IY.
Therefore, although the cause is unknown, the laser source used for IY's PF Oplev was found to have a significant power loss.
We plan to contact THORLABS later to inquire about the cause of the reduced output and whether it is possible to repair the light source that has gone bad.
1. I installed a seismograph on ITMX length oplev. (Fig 1)
2.We recalibrated the MN length oplev and PF length oplev in ITMX.As a result of readjustment, the Pitch signal no longer appears in LEN VER. (Fig 2 & Fig 3)
3.I have investigated the relationship between the position of the beam spot on the QPD and the effect of noise on OpLev.Measurements were made using ITMX PF length oplev and MN length oplev and varying HOR values of 0, 100, and 200.After the measurement was completed, I performed the length oplev QPD centering.
Seismometer information:
(CH1)X axis direction
K1:PEM-PORTABLE_IYC_RACK_IY0_ADC0_DSUB5_OUT_DQ
(CH2)Y axis direction
K1:PEM-PORTABLE_IYC_RACK_IY0_ADC0_DSUB6_OUT_DQ
(CH3)Z axis direction
K1:PEM-PORTABLE_IYC_RACK_IY0_ADC0_DSUB7_OUT_DQ
Fig.1. showed the spectrum for installed seismometer and BS/IXV seismometers.
When we adjusted the position of MN L QPD and PF L QPD, we excited.
0.35 Hz MNP 10 cnt (For MN L QPD)
1.62 Hz BFL 10000 cnt (For PF L QPD)
We removed the accelerometers that had been installed to investigate and characterize the length sensing oplevs for MN and PF stage.
Then, we performed the centering of MN/PF oplevs.
I found a questionable and interesting glitch (the channel is K1:VIS-MCI_TM_WIT_P_DQ) in the middle of night.
It seems that it wasn’t caused by human or earthquake.
[Details]
It has been known for a long time that glitches appear in MC oplevs, but the causes and trend of these glitches have not been identified yet. So, I decided to investigate them.
The date this glitch occurred is around 2021-09-26 16:55 UTC. This glitch has three interesting points.
First, there is a jump between before and after the glitch as shown around t = 7h in Fig1. The average of wave form is clearly changed.
Second, investigating the signals of various seismometers, it seems that it wasn’t caused by earthquake because there aren’t the glitches below 1Hz like earthquake, but above 1Hz in any seismometers as shown in Figs2-5 which represent the classified bands RMS from each seismometer (The Figs are Fig2: EXV, Fig3: EYV, Fig4: IXV, Fig5: MCF.). And then, these seismometers are separated each other. But nevertheless, the glitches in these signals occurred simultaneously. It is interesting.
Finally, it occurred in the middle of night, so it seems that it wasn’t caused by human.
So, it is worth investigating more. I want to advance the investigation, for example investigating other channels at same time, analyzing on Time-Frequency plane or Q-transform.
