jeffrey.kissel@LIGO.ORG - posted 22:08, Monday 05 June 2017 - last comment - 07:56, Tuesday 06 June 2017(36670)
What happened to the CARM/IMC Gain? Part 2.
J. Driggers, S. Dwyer, J. Kissel We were finally able to get past the CARM reduction and recover some semblance of nominal low noise this evening (in that ISC_LOCK guardian state, but running on SRC1 / SRM on dither alignment, having not reduced the 9 MHz modulation depth, beam diverters not closed; call this "abnormal low noise"). We've been able to get there a few times (no clue yet on why the CARM reduction has started to work). Having done so, we measured the CARM loop in a few configurations: (1) 2.0W input power, DC readout (literally, the ISC_LOCK state DC_READOUT), with the IMC boost disengaged (2) 29.5W input power, abnormal low noise, (without NOISE_TUNINGS so) with no IMC boost, just arrived at high power (3) 29.5W input power, abnormal low noise, (without NOISE_TUNINGS so) with no IMC boost, after the IFO has cooked for ~10 mins (4) 29.5W input power, abnormal low noise, (without NOISE_TUNINGS so) with no IMC boost, after the IFO has cooked for ~20 mins (5) 29.5W input power, abnormal low noise, with IMC boost, after the IFO has cooked for ~20 mins Attached are the results, and below are the observations: (1) UGF at 18.8 kHz (2) 12.3 kHz (3) 7.97 kHz (4) 6.36 kHz (5) 6.46 kHz There is no place where there is less phase margin than 45 deg. Here's *a* supposition: - Once we get up to high power, we're losing CARM plant optical gain. CARM plant optical gain is defined by the beat of the 9 MHz modulation and the carrier's "audio" frequency sidebands. While 9MHz is fine because it doesn't enter the IFO, we are losing REFL carrier (and corresponding audio sidebands) as the IFO heats up. This is because we're in an alignment / loss space in which we're closer to the critical coupling point, where all carrier power gets sucked up (and lost in) the IFO. That means there's less CARM signal side bands, which means lower CARM gain. We correspondingly / corroboratively have a lower recycling gain (29, where it used 31). See previous aLOG about critical coupling point: LHO aLOG 29474. Measurement Notes: (1) 30 mVpk excitation for all above measurements (2) still captured on ASCII because GPIB isn't working.
Non-image files attached to this report
Comments related to this report
The critical coupling point is irrelevant for the REFL optical gain. The error signal is carrier light in the orthogonal phase produced by the arms, which then beats against the 9 MHz sidebands. The amount of DC light on the REFL port only changed the shot noise. The real culprit is most likely the 9 MHz sidebands which are affected by the point absorber. In the past, we lost roughly a factor of 2 in the 9 MHz build-up—indicating significant losses in the PRC.