Sheila, Elli, Evan, Ryan, Lisa Apparently today was a particularly bad day for the whole locking business. To get the DRMI locking with the arms off-resonant, we had to retune the locking gains as previous settings were not working, meaning we could not get a single lock of DRMI + off resonant arms in ~ 20 min. After carefully aligning the DRMI (with no arms), we changed the locking settings as follows: DRMI ---> DRMI with off-resonant arms MICH 3 ---> 3 PRCL 15 ---> 8 (it was 11) SRCL -50 ---> -40 (it was -45) It would be extremely helpful to test these new settings again several times, and do some systematic lock acquisition studies from DRMI to DRMI with off-resonant arms in the morning. It still takes too long to acquire lock. We then went through the locking sequence, and stop at some point to give Evan and Elli a chance to measure the opto-mecahnical TF of the DHARD alignment DOF, so we could design a good filter to increase the bandwidth of DHARD, if necessary. We broke the lock in the process, and Sheila struggled to get ALS DIFF working right after that. While investigating the problem, Ryan found out that we could not get any actuation at all from the ETX L1 stage. Please investigate/fix this in the morning, a manual reboot at the end station might be required. On a positive side: * Evan and Elli got a good measurement of the DHARD plant and designed a compensation filter for that, so we should be ready to test it; * ETMY L1 has now a working L2P filter, so we could eventually try DARM actuation to both end masses, if necessary.
What new ETMY L2P are you referring to?? We found no filters engaged and a gain of 1.0, and the output OFF.
The tidal feedback goes to L1 state. Locking the Y-arm on green and engaging the slow servo caused a really bad L2P with the above configuration. We did a burt restore to yesterday morning (L1 L2P FM8 z0.5p10 gain -1, output ON), and everything is fine now.
The EY L1 L2P state that I found this morning is not what Ryan designed last night ( FM9 (zpk([],[0.5],1,"n")gain(0.001)), Gain 1 and output ON). However, this does not work with the slow servo on.
The EX L1 L2P FM9 was installed zpk([],[0.5],1,"n")gain(0.001). The overall filter gain remained 5.2. Not much invesitgation has been done, but based on the green transmission dips, it seems that the old FM10(-60dB) setting is better.
Elli, Evan
We have taken a TF for the DHARD pitch plant. Now filters are installed in ETMX/Y L2 LOCK P which invert this plant.
With the DHARD loop open (and oplev damping engaged), we drove band-limited white noise (up to ≈4 Hz) into DHARD EXC, which drives the ETM PUMs differentially in pitch. We then measured the transfer function DHARD EXC → DHARD IN1. This gives the plant transfer function for the DHARD loop, which we then fit by eye in Matlab to a zpk model.
The fit is not so great in phase above 0.5 Hz, but maybe as a first try it is good enough to allow us to bump up the DHARD bandwidth anyway.
The filter modules are FM6 and FM7.