Actuator efficiencies of BS (klog14265) and PRM (klog13020)
- BS_act: 5.46e-12 m/ct
- PRM_act: 1.0e-9 m/ct
Optical gains computed as
{MICH1, PRCL1}_IN1_DQ f^2
----------------------- * ----------------
{MICH2, PRCL2}_OUT_DQ {BS, PRM}_act
- BS_opt: 4.38e8 ct/m
- PRM_opt: 1.5e11 ct/m
I have reconstructed PRCL and MICH displacement sensitivity on Mar 26, 2020 offline using the calibration factors in klog #14516.
Optical gain at that time might be different, but MICH sensitivity was ~4e-15 m/rtHz and PRCL sensitivity was ~2e-15 m/rtHz at 100 Hz.
The spectrum looks reasonable.
What I did:
1. Got the following channel data from
K1:CAL-CS_PROC_(MICH|PRCL)_DISPLACEMENT_DQ
K1:LSC-(MICH|PRCL)_IN1_DQ
K1:LSC-(MICH|PRCL)_OUT_DQ
during GPS time 1269260118-1269260318. The spectrum data was taken using Pastavi.
2. Reconstructed the displacement sensitivity using the calibration factors in K1:CAL-CS_PROC_(MICH|PRCL)_FILT_TM and K1:CAL-CS_PROC_(MICH|PRCL)_FILT_INVC at that time to check the online reconstruction.
3. Reconstructed the displacement sensitivity using the calibration factors in klog #14516 to reconstruct offline. The displacement sensitivity is simply reconstructed in spectrum with:
[Displacement sensitivity] = sqrt([calibrated error spectrum]**2+[calibrated feedback spectrum]**2)
Result:
- Online calibration was totally wrong. PRCL error signal didn't have any calibration factor. MICH calibration factor was the old one. MICH had three stage 10 Hz - 100 Hz whitening filter.
- Offline calibration seems reasonable considering MICH UGF of 8 Hz and PRCL UGF of 13 Hz.
Some notes to remember:
- BS actuation efficiency in klog #14516 is 5.46e-12 m/cnt, while that in klog #13051 is 6e-11 m/cnt (which was used in online calibration). Why inconsistent?
- MICH is differential BS to ITM length (not BS longitudinal motion).
Next:
- Noise budget MICH and PRCL
- Calibrate CARM
4/21 measurement
optical gain = 15.7 ct * 4 * pi / 1064e-9
= 1.85e8 ct/m (from displacement to POP17Q_ERR)
C_mich = optical gain * input matrix / laser power norm
= 1.85e8 * 0.6 / 2.48
= 4.48e7 ct/m (from displacement to MICH_IN1)
(A_bs * C_mich) is shown in figure 1.
A_bs = (A_bs * C_mich) / C_mich
= 5.63e-11/f^2 m/ct (from BS_ISCINF_L_OUT to displacement)
Today's measurement
optical gain = 13.0 ct * 4 * pi / 1064e-9
= 1.54e8 ct/m (from displacement to POP17Q_ERR)
C_mich = optical gain * input matrix / laser power norm
= 1.54e8 * 1.0 / 2.64
= 5.83e7 ct/m
(A_bs * C_mich) is shown in figure 2.
A_bs = (A_bs * C_mich) / C_mich
= 6.28e-11/f^2 m/ct (from BS_ISCINF_L_OUT to displacement)
These values are consistent with Ogaki-kun's value (klog11754) and Kokeyama-san's value (klog14548).
In rough estimation, the largest error is in the estimation of the optical gain (~2 ct = ~15%). Coherence of the transfer function measurement is almost 1.0.
As Kokeyama-san mentioned in klog14548, maybe klog14265 is typo.
I have updated the offline reconstruction to use BS actuator efficiency of 5.63e-11/f^2 m/ct and optical gain of 4.48e7 ct/m, which is reported in klog #14554 as April 21 value.
MICH sensitivity at low frequencies got worse and the cross-over frequency between the calibrated feedback signal and calibrated error signal looks a bit too high compared with MICH UGF.
BS actuator efficiency got x10 higher (5.46e-12 m/ct -> 5.63e-11 m/ct) and optical gain got x10 higher (4.38e8 ct/m -> 4.48e7 ct/m) from the previous offline reconstruction. Cross-over frequency look fine.
MICH sensitivity was ~4e-14 m/rtHz at 100 Hz, which sounds too bad compared with PRCL.
4/21 measurement
optical gain = 15.7 ct * 4 * pi / 1064e-9
= 1.85e8 ct/m (from displacement to POP17Q_ERR)
C_mich = optical gain * input matrix / laser power norm
= 1.85e8 * 0.6 / 2.48
= 4.48e7 ct/m (from displacement to MICH_IN1)
(A_bs * C_mich) is shown in figure 1.
A_bs = (A_bs * C_mich) / C_mich
= 5.63e-11/f^2 m/ct (from BS_ISCINF_L_OUT to displacement)
Today's measurement
optical gain = 13.0 ct * 4 * pi / 1064e-9
= 1.54e8 ct/m (from displacement to POP17Q_ERR)
C_mich = optical gain * input matrix / laser power norm
= 1.54e8 * 1.0 / 2.64
= 5.83e7 ct/m
(A_bs * C_mich) is shown in figure 2.
A_bs = (A_bs * C_mich) / C_mich
= 6.28e-11/f^2 m/ct (from BS_ISCINF_L_OUT to displacement)
These values are consistent with Ogaki-kun's value (klog11754) and Kokeyama-san's value (klog14548).
In rough estimation, the largest error is in the estimation of the optical gain (~2 ct = ~15%). Coherence of the transfer function measurement is almost 1.0.
As Kokeyama-san mentioned in klog14548, maybe klog14265 is typo.