The aim of the shift was to perform a scan of the MICH_SET to see if that influences the optical spring effect we see in the MICH2DARM coupling, which differs from the optical spring in the DARM transfer function and also to test a new CARM2DARM Subtraction filter.
First we performed the following injections and MICH_SET changes:
- Noise injection at MICH_SET = 0, GPS time: 1418484683
- Move to MICH_SET = 2 at 15:50:50 UTC
- Noise injection at MICH_SET = 2, GPS time: 1418486005
- Move to MICH_SET = 6 at 16:05:10 UTC
- Noise injection at MICH_SET = 6, GPS time: 1418486951
- Move to MICH_SET = 14 at 16:26:50 UTC
- Noise injection at MICH_SET = 14, GPS time: 1418488135
- Move to MICH_SET = 26 at 16:42:35 UTC
- Noise injection at MICH_SET = 26, GPS time: 1418489335
- Move to MICH_SET = -6 at 17:17:10 UTC
- Noise injection at MICH_SET = -6, GPS time: 1418491637
- Move to MICH_SET = -14 at 17:38:00 UTC
- Noise injection at MICH_SET = 0, GPS time: 1418492441
- Move to MICH_SET = -26 at 17:49:10 UTC
Then the interferometer unlocked when we were close to MICH_SET = -26. We haven't been able to determine whether the MICH offset was the cause or if there was another problem.
In Fig. 1 the DARM and MICH2DARM transfer functions are shown for the different offsets. It is clear that the MICH_SET increases the OS frequency for higher offsets (14, 26). It is also difficult to judge whether the effect is symmetrical. The total offset has to be considered since a the loop for the RIN also adds an offset which is not included in MICH_SET (but it is in MICH_SET_TOT). This offset was roughly 4 and when adding this to the MICH_SET values in the plot, the -14 point and +6 point are -10 and +10 in total MICH offset, and a small change in the OS seems to be happening which is roughly similair for both. The measurement at -26 MICH_SET would have helped to confirm the symmetry, but since we only had half an hour left before the end of the shift and we weren't sure whether the high MICH_SET caused the unlock, we decided against measuring this point.
We were mostly interested in determining whether the MICH_SET affected the OS effect in MICH2DARM but not in DARM transfer function to see if this was the cause for the fluctuating alpha filter. In Fig. 2 the alpha filter for the different MICH_SET is shown. Particularly in the phase behaviour it is clear that the MICH_SET indeed affects the alpha filter and particularly the differential optical spring effect between the DARM and MICH2DARM transfer function, as the phase behaviour shows a clear increase in phase from the -14 offset to the +26 offset, also indicating that this effect is not symmetrical around a zero offset. Further analysis and interpretation of the data is still needed.
After the unlock we moved to the CARM2DARM Subtraction filter. The subtraction had been left off since it wasn't working properly. After analyzing the latest injection, a new filter was fitted (see Figure. 3) on the most recent injection (december 2nd). We implemented this and the result is shown in Fig. 4, with purple no subtraction, red subtraction with gain = 1, and blue the subtraction with gain = -1. After flipping the sign we managed to get roughly an order of reduction in CARM noise in DARM, so we left the filter on with the gain = -1.
After we concluded the work we left the interferometer in science mode.
