Reports until 01:58, Thursday 13 December 2018
H1 ISC
sheila.dwyer@LIGO.ORG - posted 01:58, Thursday 13 December 2018 - last comment - 10:51, Thursday 13 December 2018(45893)
locking tonight

Some notes on locking tonight (Rana, Craig, Dan Brown, Terra, Sebastian, Sheila, Hang, TVo, Danny Vander-Hyde)

Violin modes:

We tried Jenne's new state which turns on all the ASC at the same time, which has worked twice.  We have increased the thresholds used by the convergence checker a lot. 

We have also turned on the OMC DCPD low pass filter.  The OMC guardian can now be used to add and remove the low pass filter while we are locked on DC readout, and to add one additional stage of whitening, although we haven't tested the additional stage of whitening while on DC readout. The low pass is now turned on when we lock the OMC, which will make our dark noise bad enough to add noise to DARM.  This should help us to avoid violin mode problems while we power up, we can either remove the low pass or add the next whitening stage to recover the sensitivty. 

We have had lots of trouble with ALS, it seems like we have fast glitches which might be a problem with the fiber.  We also have the micro seism around 1um/sec and wind gusts up to 40mph, so that is making the locking more difficult.

DRMI locking has continued to be much better with the preloading reduced.

Comments related to this report
craig.cahillane@LIGO.ORG - 05:20, Thursday 13 December 2018 (45899)
Link 

When we acquire DRMI (ISC_LOCK_STATE_N ~100), the PRM is badly yawed.  
We've lost lock tonight during ASC a couple of times, so I took Jenne's new slow route through it this time.
On successful locks tonight, the PRM (PRC1) loop has been railed slowly bringing the PRM into a reasonable place.  Otherwise, our POPAIR18 has been incredibly "dippy", AKA large sideband power swings, often leading to locklosses after acquiring DRMI.
We need to engage PRC1 right after catching DRMI to bring the PRM into a reasonable alignment before going on.
Images attached to this comment
terra.hardwick@LIGO.ORG - 05:48, Thursday 13 December 2018 (45900)
Link

I attach the simulated HOM spacing of the arms: you can see the rises from self heating after we power up and then we lose lock at about the same thermal state/arm cavity config each time.

Images attached to this comment
thomas.vo@LIGO.ORG - 10:51, Thursday 13 December 2018 (45905)
Link

I attached a screen comparing the TCS with 50 Watt pre-loading (left) vs 30 Watt pre-loading (right). 

We were able to stay locked at 25 watts input for many hours with the 50 watt TCS pre-loading but with about 700 mW of CO2 power left over on both CO2X and CO2Y.  The 30 Watt TCS settings seem to be unstable within a few minutes of going up to 25 Watts of PSL input power.  Both of them display a drop in POPAIR_B_RF18 that seems consistent with a thermal time constant and drop lock at around 53-54 counts. So maybe this points to more room for improvement with annular heating to try and push for higher stable power operation, but not so much pre-loading that it makes DRMI difficult to lock.

EDIT:

Sheila showed me some data from Nov 9th just before the vent where there was a 27 Watt input lock for about 4 hours so it is possible with the 30 Watt TCS pre-loading settings to stay higher.

Images attached to this comment