At 01:27 (18:27) got a CRC error due to problem with H1IOPISCEY. Called Dave B., see aLOG 50009. Lost lock at 01:53 (18:53) as expected when Dave rebooted H1ISCEY.
Reset End-Y ISI and HEPI tripped WDs.
At 02:01 (19:01) started relocking. Initially had some issues getting Y-Arm Green to lock. At the time there were several wind gusts in the mid 30mph range. After things settled a bit were able to reliably lock both arms green. Had to tweak the BS and PRM to get through DRMI_1F. After this no additional adjustments were needed.
At 03:02 (20:02) and 03:23 (20:23) lost lock at ENG_SOFT_LOOPS. Georgia is looking into these lock losses.
Have been seeing 30 plus mph gusts at End-X, which is not helping the relocking.
We lost lock engaging the soft loops a few times this evening.
It looked like the corner ASC could not keep up when we switched on the arm dither loops, symptom: power recycling gain goes up, rf18 buildup drops, we lose lock.
I removed 10dB of gain from the arm dither loops (pit4, pit5, yaw4, yaw5) when they first come on (set in PREP_ASC_FOR_FULL_IFO), and added the 10dB back in later in ENGAGE SOFT LOOPS. This worked by hand, so I put it in the guardian.
Dave was thinking we could restart this ISCEY computer without affecting the SEI or the SUS platforms and we all agreed until we realized we shouldn't have. There was no impact on the models of the SEI or the SUS but, ALS is bleeding the Tidal Drive from the ETM to the HEPI. When the ALS came back on with zero Tidal Drive, the HEPI was jerked back to zero offset. The HEPI tripped on this 6 seconds before the ISI tripped when the T240 tilted off or got slung sideways by the HEPI. Attached is 4 hours showing the tidal drive on HEPI coming from the ALS (lower left) going flat and then jerking this way and that on its way to zero when the other channels, the ISI's T240s, start railing.
I find interesting that the T240 signals all go flat when the ALS does with the Kernel panic but maybe that would be expected too.
We'll endeavor to implement some procedure to hold this output on the HEPI and then bleed it off at an acceptable rate.