I tried to calibrate the MCE actuator based on the FSR of IMC.
After then, actuator efficiency ratio between MCE and NPRO PZT was estimated by using some TF measurements.
- MCE actuator efficiency: 440 ct/FSR ~ 12.8/f^2 kHz/ct (ct@K1:VIS-MCE_TM_TEST_L_OUT) ~ 41.9 MHz/V (V@DAC)
- PZT efficiency: 5.56 MHz/V (V@K1:IMC-SERVO_NPRO_TEMP_BIAS_IN2)
BTW, past measurements are as follows.
A_mce
- klog#2985: 17.2 MHz/V
A_pzt
- klog#2853: 1.48 MHz/V
- klog#13012: 3 MHz/V
Hmm...
-----
Measuring MCE actuator efficiency by FSR.
1. Unlocked IMC
2. Moved MCE via K1:VIS-MCE_TM_TEST_L_OFFSET
(I changed from -500ct to 1500ct)
3. Read a change on offset count from the 1st resonance to next resonance.
It seems that 1FSR is equivalent to roughly 440ct. (See Fig.1)
But the relationship between the resonance point and the absolute amount of offset count does not seem to be very repeatable.
(It probably comes from the drift of laser frequency and natural motion of suspensions.)
So we need to consider how to evaluate the systematic error of this measurement.
Measuring actuator efficiency ratio and NPRO PZT efficiency
1. Locked IMC
2. Measured TF1 from K1:VIS-MCE_TM_TEST_L_EXC to K1:IMC-SERVO_OUT2_OUT [= A_mce * C_imc / (1 + G_imc)] (See Fig.2)
3. Measured TF2 from K1:IMC-SERVO_NPRO_TEMP_BIAS_EXC to K1:IMC-SERVO_OUT2_OUT [= - A_temp * C_imc / (1 + G_imc)] (See Fig.3)
4. Measured TF3 from K1:IMC-SERVO_NPRO_TEMP_BIAS_OUT to K1:IMC-SERVO_SLOW_DAQ_OUT [~= - A_temp / A_pzt] (See Fig.3)
5. Computed A_pzt/A_mce = TF1 / TF2 / TF3 ~= 435 * f^2 ct/V (See Fig.4)
Finally, we can obtained A_pzt ~= 5.56 MHz/V
I cannot use K1:IMC-SERVO_MIXER_DAQ_OUT instead of K1:IMC-SERVO_OUT2_OUT because large gain is applied on universal whitening filter.
