[Sheila, Keita, Jenne, Patrick]
We have been struggling to lock anything at all today. Sheila is going to post about the loop measurements that we made today, but this alog is a cry for help with regard to the ISS.
The ISS average diffracted power has been much more noisy than usual, ever since Friday. Since today's PSL work, it's even worse. We suspect that this may be bad enough that we are unable to lock. We can no longer keep the integrator for the first ISS loop engaged for more than a few minutes before the loop starts to oscillate. But, with or without the integrator on, the diffracted power is moving around like crazy.
At this point, we are leaving the IFO in Down, and hopefully someone from Team PSL will come in and look at the ISS tomorrow, and then maybe locking will be possible.
As a side note, it appears that everything in the PSL is misaligned by a bit. Both the PMC and RefCav transmissions are about 25% lower than normal, with corresponding increases in their reflected powers. This is probably the reason that we've had to increase the FSS loop gain earlier today to make locking even the IMC possible (see Sheila's alog).
Attached are measruements of the IMC open loop gain. We cannot measure the FSS gain without going into the PSL, but Peter King posted some measurements from May
The first attached plot shows that the IMC goes unstable around 200 kHz when the FSS gain drops by only 2dB. Since the power transmitted by the reference cavity droped 25% after the laser problem, this probably explains why we were having trouble keeping the IMC locked with our old FSS gain of 14dB. (see Hoacun's measurement of the time when we reduced the fast gain from 20 to 14 dB for a comparison.) It seems like we want more than 2dB of gain margin though... The orange trace shows that the IMC OLG continues to improve as we turn the common gain up to 20dB, but in this configuration the FSS cannot recover from a lockloss. If this is the reason we keep the FSS gain marginal, we could write a gaurdian to automate relocking at a lower gain and turning up the gain once locked.
We checked that the IMC optical gain has not changed since Evan H's mid June measurement. (taken on the day the IMC gain was reduced by 4 dB from a IN1 gain of 20dB for 2 Watts to 16 dB at 2Watts which gives us a UGF of about 65kHz).
The second plot shows what happens when we change the fast gain by +/- 4 dB, reducing the gain makes gain peaking at 200 kHz.
The data attached is for the configuration we are leaving, IMC locked at 2 Watts with 16dB IN1 gain, FSS common gain 16 dB, FSS fast gain of 22 dB.
Also, there seems to be a problem with ALS DIFF after today's maintence. We tried flipping the bias sign back, but that didn't help. Since we still have laser/FSS/IMC problems we didn't investigate much further.
My guess based on the LLO experience of late is its the ISS AOM. My suggestion (from afar) is to realign and recalibrate the ISS AOM (its something that Christina B from AEI suggested should be done each time laser alignment could change). Since Ive done ours, our fluctuations that were on the order of %'s during O1 and before has really settled down to barely vary at all (hopefully I haven't jinxed myself now).
However doing this to the ISS AOM will probably misalign the PMC a bit (but if its already misaligned and need to redo the alignment anyway, now is maybe a good time to try realigning and recalibrating the ISS AOM)