craig.cahillane@LIGO.ORG - posted 00:18, Wednesday 24 April 2019 - last comment - 17:40, Thursday 25 April 2019(48717)
Frequency Noise Measurements today
Today I took some frequency noise measurements. Posted are the results and the file locations.
I increased CMB IN1 from +9dB to +12dB. We should always do this after thermalization of the IFO.
We can't increase the CARM digital gain too much or we'll hit the FSR with the CARM UGF. We have to wait for the optical gain to decay and replace it with digital gain.
POP18 NORM MON is at ~47 cts now, but is at around 60 when we first lock. We should wait until we hit around 50 cts, then up the CARM gain.
IMC Locked Alone at 35 W
- IMC REFL stitched spectra : /ligo/home/craig.cahillane/Git/IFO/IMC/data/Spectra/StitchedSpectrum_20190423_IMC_REFL_IMON_Spectra_IMCAlone_35WInput_IN1_22dB_SecondBoostOn.txt
- IMC OLG : /ligo/home/craig.cahillane/Git/IFO/IMC/data/TFs/20190423_IMC_OLG_IMCAlone_35WInput_IN1_22dB_SecondBoostOn.txt
- IMC MCL Crossover : /ligo/home/craig.cahillane/Git/IFO/FrequencyNoise/data/20190423_IMC_MCL_Crossover.xml
- MC2 to IMCF TF (Refs 16 and 17): /ligo/home/craig.cahillane/Git/IFO/FrequencyNoise/data/20190423_MC2_to_IMCF_35W.xml
Full Lock at 35 W
- REFL B Spectra : /ligo/home/craig.cahillane/Git/IFO/CARM/data/Spectra/StitchedSpectrum_20190423_REFL_B_35W_Input_CMBIN1Gain_9dB.txt
- REFL A Spectra : /ligo/home/craig.cahillane/Git/IFO/CARM/data/Spectra/StitchedSpectrum_20190423_REFL_A_35W_Input_CMBIN1Gain_9dB.txt
- CARM OUT2 Spectra : /ligo/home/craig.cahillane/Git/IFO/CARM/data/Spectra/StitchedSpectrum_20190423_CARM_OUT2_35W_Input_CMBIN1Gain_9dB.txt
- IMC REFL Spectra : /ligo/home/craig.cahillane/Git/IFO/CARM/data/Spectra/StitchedSpectrum_20190423_IMC_REFL_35W_Input_CMBIN1Gain_9dB.txt
- IMC TEST1 Spectra : /ligo/home/craig.cahillane/Git/IFO/CARM/data/Spectra/StitchedSpectrum_20190423_IMC_TEST1_35W_Input_CMBIN1Gain_9dB.txt
- CARM OLG w/ CMB IN1 = +9dB : /ligo/home/craig.cahillane/Git/IFO/CARM/data/TFs/20190423_213400_20190423_CARM_OLG_FullLock_35W_Input_60mV_Exc.txt
- CARM OLG w/ CMB IN1 = +12dB : /ligo/home/craig.cahillane/Git/IFO/CARM/data/TFs/20190423_214036_20190423_CARM_OLG_FullLock_35W_Input_60mV_Exc_12dB_CMBIN1Gain.txt
- IMC OLG w/ CMB IN1 = +9dB : /ligo/home/craig.cahillane/Git/IFO/IMC/data/TFs/20190423_212909_IMC_OLG_FullLock_35W_Input_0dBm_Exc.txt
- IMC OLG w/ CMB IN1 = +12dB : /ligo/home/craig.cahillane/Git/IFO/IMC/data/TFs/20190423_214142_IMC_OLG_FullLock_35W_Input_0dBm_Exc_12dB_CMBIN1Gain.txt
- LSC MCL Crossover : /ligo/home/craig.cahillane/Git/IFO/FrequencyNoise/data/20190423_LSC_MCL_Crossover.xml
- MC2 to REFL9 Cal TF (Current): /ligo/home/craig.cahillane/Git/IFO/FrequencyNoise/data/20190423_MC2_to_IMCF_35W.xml (This is the same as the MC2 to IMCF template above. The current references are in full lock, Refs 16 and 17 are from IMC locked alone.)
- Frequency Noise Injs +9dB: /ligo/home/controls/craig.cahillane/Git/IFO/FrequencyNoise/data/Injections/20190423/1240118413_GPSstart_FrequencyNoise_CMB_EXC_inj_2000_7000_Hz.pkl
/ligo/home/controls/craig.cahillane/Git/IFO/FrequencyNoise/data/Injections/20190423/1240118674_GPSstart_FrequencyNoise_CMB_EXC_inj_600_2000_Hz.pkl
/ligo/home/controls/craig.cahillane/Git/IFO/FrequencyNoise/data/Injections/20190423/1240118847_GPSstart_FrequencyNoise_CMB_EXC_inj_175_600_Hz.pkl
/ligo/home/controls/craig.cahillane/Git/IFO/FrequencyNoise/data/Injections/20190423/1240119020_GPSstart_FrequencyNoise_CMB_EXC_inj_50_175_Hz.pkl
/ligo/home/controls/craig.cahillane/Git/IFO/FrequencyNoise/data/Injections/20190423/1240119191_GPSstart_FrequencyNoise_CMB_EXC_inj_15_50_Hz.pkl
- Frequency Noise Injs +12dB: /ligo/home/controls/craig.cahillane/Git/IFO/FrequencyNoise/data/Injections/20190423/1240119574_GPSstart_FrequencyNoise_CMB_EXC_inj_2000_7000_Hz.pkl
/ligo/home/controls/craig.cahillane/Git/IFO/FrequencyNoise/data/Injections/20190423/1240119754_GPSstart_FrequencyNoise_CMB_EXC_inj_600_2000_Hz.pkl
/ligo/home/controls/craig.cahillane/Git/IFO/FrequencyNoise/data/Injections/20190423/1240119927_GPSstart_FrequencyNoise_CMB_EXC_inj_175_600_Hz.pkl
/ligo/home/controls/craig.cahillane/Git/IFO/FrequencyNoise/data/Injections/20190423/1240120100_GPSstart_FrequencyNoise_CMB_EXC_inj_50_175_Hz.pkl
/ligo/home/controls/craig.cahillane/Git/IFO/FrequencyNoise/data/Injections/20190423/1240120272_GPSstart_FrequencyNoise_CMB_EXC_inj_15_50_Hz.pkl
- LSC MCL Noise Injection +9dB: /ligo/home/controls/craig.cahillane/Git/IFO/MCL/data/Injections/20190423/1240123470_GPSstart_MCL_inj_5_200_Hz.pkl
- LSC MCL Noise Injection +12dB: /ligo/home/controls/craig.cahillane/Git/IFO/MCL/data/Injections/20190423/1240120857_GPSstart_MCL_inj_5_200_Hz.pkl
- PRCL Noise Injection +9dB: /ligo/home/controls/craig.cahillane/Git/IFO/PRCL/data/Injections/20190423/1240123215_GPSstart_PRCL_inj_5_200_Hz.pkl
- PRCL Noise Injection +12dB: /ligo/home/controls/craig.cahillane/Git/IFO/PRCL/data/Injections/20190423/1240121461_GPSstart_PRCL_inj_5_200_Hz.pkl
- Intensity Noise Inj +9dB: /ligo/home/controls/craig.cahillane/Git/IFO/IntensityNoise/data/Injections/20190423/1240122973_GPSstart_Intensity_inj_10_7300_Hz.pkl
- Intensity Noise Inj +12dB: /ligo/home/controls/craig.cahillane/Git/IFO/IntensityNoise/data/Injections/20190423/1240122556_GPSstart_Intensity_inj_10_7300_Hz.pkl
I upped the CMB IN1 gain from 9 to 12 dB for a couple of CARM and IMC OLGs. From the CARM OLG (PDF 3), we seem to have about 3 dB of clearance from the FSR with CMB IN1 = +12dB.
We know that frequency noise is starting to limit us with our squeezing levels.
Images attached to this report
Non-image files attached to this report
Comments related to this report
Quick comparison of the frequency noise injections with low (CMB IN1 +9dB) and high (CMB IN1 +12 dB) CARM gain. PDF 1 shows four ASDs: 1) DARM during a 2 to 7 kHz frequency noise injection into the common mode board, with +9dB on the CMB IN1 gain slider. 2) DARM during a 2 to 7 kHz frequency noise injection into the common mode board, with +12dB on the CMB IN1 gain slider. 3) Nominal DARM 4) Frequency noise projection into DARM for +9dB The frequency noise levels apparent in DARM decreased when the CARM analog gain was increased. This is because we have squashed the frequency noise imposed by the IMC shot noise. REFL B with high CARM gain was not measured yesterday. The CARM to DARM coupling TF did not change between gain changes. This is expected.
Non-image files attached to this comment
Some long term questions to answer: - Unmodeled CARM OLG hump at 18 kHz -- Daniel claims this is 9 MHz resonating in the arms, we can try to model this -- Was not apparent in Evan's thesis Figure 2.7 -- Not that apparent at Livingston (LLO alog 37620) - Make sure OMC control noise not limiting DARM at current frequency noise levels -- Georgia made OMC controls projections in the current noise budget at 30 W, showed noise way below DARM. -- Controls noise not linear with frequency noise (alog 45768) - Model shot noises for CARM, IMC -- Can get a quick win of sqrt(2) from using both REFL detectors -- Cyclostationary noise on REFL -- Should increase optical gain of IMC to squash its shot noise so it doesn't appear in DARM --- Increase modulation depth for IMC --- Add fast shutter and rotation stage to IOT2L to control power levels on IMC REFL
Attaching a screenshot shows DARM with the two LSC-REFL_SERVO_IN1 gains (grey = 12dB, red = 9dB). The improved frequency noise suppression is visible in DARM above 3.5 kHz, and also in the squeezer BLRMS (bottom time series), which looks at DARM at 4.68 kHz.
Images attached to this comment
Additional CARM Spectra from 0.5 Hz to 5 MHz with the new changed configurations: CMB IN1 = +12 dB: REFL B : /ligo/home/craig.cahillane/Git/IFO/CARM/data/Spectra/StitchedSpectrum_20190425_REFL_B_35W_Input_CMBIN1_12dB.txt REFL A : /ligo/home/craig.cahillane/Git/IFO/CARM/data/Spectra/StitchedSpectrum_20190425_REFL_A_Spectrum_35W_Input_CMBIN1_12dB.txt CARM OUT2 : /ligo/home/craig.cahillane/Git/IFO/CARM/data/Spectra/StitchedSpectrum_20190425_CARM_OUT2_Spectrum_35W_Input_CMBIN1_12dB.txt IMC REFL : /ligo/home/craig.cahillane/Git/IFO/CARM/data/Spectra/StitchedSpectrum_20190425_IMC_REFL_Spectrum_35W_Input_CMBIN1_12dB.txt IMC TEST1 : /ligo/home/craig.cahillane/Git/IFO/CARM/data/Spectra/StitchedSpectrum_20190425_IMC_TEST1_Spectrum_35W_Input_CMBIN1_12dB.txt CMB IN1 = +6dB, CMB IN2 = +6dB (split control for REFL A and B) REFL B : /ligo/home/craig.cahillane/Git/IFO/CARM/data/Spectra/StitchedSpectrum_20190425_REFL_B_Spectrum_35W_Input_CMBIN1_6dB_CMBIN2_6dB.txt REFL A : /ligo/home/craig.cahillane/Git/IFO/CARM/data/Spectra/StitchedSpectrum_20190425_REFL_A_Spectrum_35W_Input_CMBIN1_6dB_CMBIN2_6dB.txt Posted are the comparison spectra of the three configurations of CARM we've been playing with:We can see that the CARM loop is gain limited at ~2kHz, since the REFL B spectrum decreased from increasing the CARM gain (Dark blue vs Light Blue). We can also see that REFL shot noise is dominating the spectrum from 3kHz down, from the switch to split sensor control (Light blue vs Orange)
+9 dB, REFL A sensor controlling CARM +12 dB, REFL A sensor controlling CARM +12 dB, Split REFL A and B sensors (+6 dB on each of the CMB Inputs)
Non-image files attached to this comment