Chris, Kiwamu, Lisa, Stefan
After the 45MHz 12.6dB boost this morning we recommissioned the interferometer:
- We realigned the whole interferometer.
- We added ND0.4 filter on REFLAIR_A to avoid saturation.
- We adjusted gains and phases to match REFLAIR_A_RF45, REFLAIR_B_RF27 and REFLAIR_B_RF135.
- First we tried approaching the fringe from the left side. We were hampered by weird demod phases in RF45, as well as the appearance of a resonance in transmission of the y-arm. This probably is the 20 mode of the 45MHz.
- We then decided to approach from the right side of the fringe. There REFLAIR_A_RF45 had the same demod phase as PRMI only.
- We acquired 700 red Hz off resonance.
- We decided to use both its I and Q for PRCL and MICH, because they looked cleaner at low frequency.
- We transitioned at 700 red Hz to RF135.
- We needed the following settings:
REFLAIR_A_RF45 147deg (same for PRMI, and with arms)
MICH GAIN 1.2 (UGF 6Hz), PRCL GAIN 0.8 (UGF 25Hz)
- Next we started walking in. We needed to update the REFLAIR_B_RF135:
73deg (PRMI),
58deg (1kHz offset),
51deg (and MICH GAIN from 1.2 to 1.0)(450Hz CARM offset)
42deg (250Hz CARM offset)
unlocked at 75Hz (up to 1 in ETMY transmission)
- Next attempt identical, but no unlock
40deg (160Hz CARM offset)
At 125Hz CARM offset we noticed that only the y-arm was building up, so we fine-tuned the DARM offset until the arms were symmetrical.
37deg (60Hz CARM offset)
- Then we marched across the river.
Some observations:
- We had ad least a factor of 10 buildup. We should get more, but we were still moving across the fringe rapidly.
- We looked at REFLAIR_A_RF9 but it was puzzling: only a large offset in I.
- We then tried a pure DARM fringe offset, but this made the 3f signals flaky.
- The 3f signals were fine across the resonance.
Resonant arms and words of wisdom.
It seems that some of the length degrees of freedom had an offset. It is not clear at this point which DOF was off from the operating point. The POP DC basically did not show a buildup at all when the arm was at a high value. The REFL DC did not become a smaller value. Plus, the demodulated signals at 1fs had an offset without any PDH-looking waveform during the arm was flashing. The attached shows the moment of the lock loss. You can see how they behaved odd.
After the people left, I made a few more attempts of this locking procedure last night. It worked robustly. The funny REFL9I offset was also seen consistently.
After the CARM offset was completely removed, I then got rid of the 135I and 135Q offsets to see if this helps the power buildup in the arm cavities. This actually made them worse --- the buildup in the arm cavities decreased by a factor 2 or so. Since the interferometer dropped the lock afterwards, I did not explore a bigger offset in the 135 signals.