Maybe we turned off the PRCL FF by changing the matrix of the
K1:LSC-OUTPUT_MTRX_3_13
from 1 to 0.
Fig.1. showed the status of several PDs when lock loss happened.
After this lock loss, the OMC DCPD B output kept about 10,000 count even no in case of the no laser.
So there are possibility to break the DCPD B by the lock loss.
We didn't check the DCPD A carefully, but this is not broken, I think.
If you have some comments or suggestion, please contact to Yokozawa.
We are sorry for inconvenient to you.
I measured OMC PDA/B spectra.
OMC PDA seems fine in terms of the noise spectrum.
Note:
I made a mistake in the legend of PDB spectrum.
Red and Blue show the 400-Ohm and 100-Ohm transimpedance measured on Jan. 31, not Jan. 24.
Consequently, the countermeasures for protecting DCPD from burning could not function under the PRFPMI+DC with 5W injection as before.
Accident reason
The strong laser beam from the resonated OMC could not be quenched quickly because of the LPF in the HV driver for OMC PZT (circuit) even if the input signal to this PZT driver was quickly terminated and swung with triangular shape signals in the DGS control system.
You can find the LPF consisting of R151(50K) and C111(0.47uF) and C of PZT(~0.5uF according to speck) to "hv_omc". The cut frequency is ~ 3.2Hz which is much lower than the frequency with which we would like to cut signals.
We need to check whether there is another possibility to slow the signal response to PZT.
One of Countermeasures
Simply, the cut of frequency of LPF should be enlarged. However, it will result in the noise introduction of the HV amp for the high-frequency ranges. Aso-kun will ask Arai-san in LIGO what is their methods.
Plan for several weeks
- Laser power of 1W at the IMC out is kept.
- Realize the stable operation of PRFPMI+RF with 10W injection beam because we encounter the yaw motion oscillation maybe due to the radiation pressure in FP cavities.
- DAC noise reduction around 100Hz.
- Total test of the DCPD protection with safe laser power
- Firstly, we measure the OMC output changes when the OMC resonance is cut by using one arm FP IR resonance and OMC lock.
- Modify the circuit if we have a reasonable plan.
- We try the same measurement and check the speed of changes of the OMC output power.
When we will repair the DCPD
- It depends on how fast the HV circuit can be modified.
- The time limit will be the end of February when the OMC black shroud will be prepared.
Several months ago I estimated the cutoff frequency of the PZT response to 1.0 Hz (klog 30535). The capacitance of the PZT you adapted seems too small; it's Noliac NAC2124-H08, its capacitance is 1380 nF.