[Stefan, Kiwamu]

The demod phases for the IMC changed again and we readjusted them.

Then after a course of adjustments and etc, we came back to the point where we can update the CARM beatnote noise spectrum.

There is some improvements at around 10 Hz but the low frequency performance was not great today probably due to seismic noise, resulting in a worse RMS.

The spectrum  update:

The attached is the new spectrum which we took tonight. Note that ISS was off tonight.

Green : our best noise curve

Blue : new spectrum with a low control BW

Red : new spectrum with a high control BW

IMC demod phase rotated again :

The demod phases for the IMC locking and WFSs rotated again. It is still unclear what caused it.  This is the second time to see such a spontaneous phase rotation. The first one just happened yesterday and the second happened this evening. In this early evening we found that the IMC locking loop lost a factor of about 3 for some reason, resulting in a lower UGF.  At this point we were OK with the demod phase which was newly set yesterday. However after a number of investigation we found that the demod phase rotated by 90 degrees backward as if it came back to the original phase.

 Before we noticed it we were simply chasing the missing gain factor. We looked at the EOM modulation source path at which we found the loose connection yesterday. The output from the distribution amplifier looked OK --- the signal at 24 MHz had 3.4 Vpk-pk. Then I entered the PSL to check the power of the same RF signal but at the end of the RF chain right before the EOM. I disconnected the SMA connector at the side panel of the EOM can and plugged it to an scope. It seemed also fine --- the RF signal had 3.11 Vpk-pk which agrees with the previous measurement done by Volker a while ago. Note that during the measurement I closed the HPO external shutter to be safe. Also I had to disconnect the SMA connector of the 45 MHz  source since this was in the way of the 24 MHz connector on the side panel. All of the SMAs were put back on after the measurement. Before getting in the PSL I switched the PSL from the science mode to the comissioning mode. I did the vice versa when existing. In addition to it Stefan was manually aligning a couple of steering mirrors on IOT2L to center the beams on length diode and both WFSs. This ended before I started messing up the PSL RF connections.

 After these investigations Stefan found that all the demod phases got rotated by approximately 90 degrees so that the IMC length signal showed up mostly in the Q-phase. So for the reason we had to start over the phase adjustment.

The new demod settings:

It seems they came back to where they were with a descrepancy of roughly 10 degrees compared with the values we had until yesterday. The below are the new settings for the WFS demod phase.

• WFS_A
  • H1:IMC-WFS_A_SEG1_PHASE_R = -185
  • H1:IMC-WFS_A_SEG2_PHASE_R = -112
  • H1:IMC-WFS_A_SEG3_PHASE_R = -185
  • H1:IMC-WFS_A_SEG4_PHASE_R = -185
• WFS_B
  • H1:IMC-WFS_B_SEG1_PHASE_R = -80
  • H1:IMC-WFS_B_SEG2_PHASE_R = -132
  • H1:IMC-WFS_B_SEG3_PHASE_R = -66.5
  • H1:IMC-WFS_B_SEG4_PHASE_R = -173

And the below is a picture of the new setting for the delay line for the length demodulation.

The IMC controls :

 With the new demod settings we became able to lock the IMC with no major problems. A good this is that the UGF came back to the original high value of  60-ish kHz, which was 25 kHz this morning due to the missing factor of 3. However the WFSs seemed slightly unstable. We saw a slow oscillation in pitch at an order of 100 mHz which was visible in the reflection CCD. We thought this was due to a too much gain even though we are not sure why. We decreased the gain of all the DOFs in the WFS by simply disabling the 10 dB gain flat filter for each of them. Note that the overall gain H1:IMC-WFS_GAIN is set to be 1. It is working fine now without an oscillation. Good.

Alignment of interferometer :

We moved on to the global alignment of the interferometer. The Y arm cavity was tweaked to maximize the green intracavity power. We touched both ITM and ETM. The BS was aligned by using the opleve as a reference. Later on we did a fine alignment of the BS by looking at the amplitude of the beatnote and maximizing it. At this point we established the green beam path so that we had a reasonably big beatnote signal at ISCT1. Then we moved on to the alignment of the infrared path. We then tweaked both IM4 and PR2 to get a high transmitted light of the Y arm in infrared. The alignment was mainly done by touching PR2 in both yaw and pitch. As a result we got 400 counts in H1:ASC-TR_A_SUM_OUT when the 00 mode is on resonance. Since the highest value we got ever was about 500 counts we concluded that this was good enough for measurement of the noise spectrum.

HEPI turned off for the night :

We found that the cavity length moving quite a lot tonight although we didn't see an apparent seismic activity in the seismic strip chart in the control room. However this was obvious since the end PDH kept dropping its lock frequently. The fluctuation was at a time scale of 1 sec or longer and hence the HEPI offloading didn't help at all as it was too fast for it. As a test we turned off the ETMY HEPI by running the script. But this tripped the HEPI and everything else (TMSY, ISI, ETMY). We tried to bring the HEPI back on by running the script but it tripped again after immediately the control were engaged. But it turned out that the cavity was quiet if the HEPI was left off. So we decided not to turn it on for tonight. We proceeded the noise measurement with the ETMY HEPI off.

A near-future task:

The IMC guardian must wait for the refcav for a long enough time when it is being recovered so that the IMC won't kick the refcav until it settles.