Lock 1:

• Some oscillations in the ASC loops at the end of power up. We survived, but I did hold us in lownoise ASC to watch the shaking damp down before moving on
  • 0.5 Hz oscillations, they showed up strongly in SRC2 P, which is usually indicative of CSOFT P. They were also visible in DHARD P
• locking with the appropriate drive align gain for ETMX L3 L2L (in guardian) removed the extra noise we saw in DARM last night
• I ran another DARM OLG measurement to check that the ETMX L3 drivealign gain was still correct and saw that it needed to be increased by 4 percent. This is now in the guardian in LSCparams
  • See attached plot. I saved this sweep (red trace) in (userapps)/lsc/h1/templates/DARM/DcDARMLVLN_reduce_bias.xml so it can be used as reference in the future
• Jenne and I increased the ADS gains to 10 to make them converge faster so we could move to camera servos
  • We eventually had to increase the Pit ADS gains to 20 to hasten the convergence
• The guardian will now set the ADS gains to 10 in lownoise ASC (used to be set to 4). Hopefully this will help the ADS converge faster so we can move to NLN with camera servos sooner.
• We did some open/close tests of the beam diverters (REFL, POP, AS) to see if there was increased scatter from the diverters being closed, but saw nothing conclusive to suggest that. It would be best to do this test with ADS lines off/using the camera servos, so the changing low frequency noise isn't so confusing
  • comparison plot from the test
• There was a lockloss when we tried to move onto camera servos (see Jenne's alog)

Lock 2:

• same oscillation in ASC loops when reaching 60W. I pushed the gain up on CSOFT P to 22 until the oscillations died down
• successful transition to camera servos
• PCAL BB taken 30 minutes into lock
• quiet time with sqz in cal meas start 23:42:36 (10 minutes long)
• new LSC injections taken now that the drivealign gain is corrected
• ETMY camera froze and we quickly returned to NLN and turned on ADS
• Gabriele and I ran another frequency noise injection using Craig's frequency noise injection caller code

3 hour mark ASC issues:

• around 3 hours into the lock, I saw a lot of motion in the soft yaw loops. Since we have had thermalization issues with CHARD Y in the past I started by dropping the gain slightly from 125, to 115 and then 110. This made the oscillation worse (it was around 3.2 Hz). The other loop I guessed was CHARD P, which also has a UGF around 3.2/3.3 Hz. Raising the gain again made the problem worse, but reducing the gain to 0.8 (20% reduction) prevented any more oscillations growing.
• CHARD P doesn't appear to be that marginal (looking at the models and measurements made by Craig it should have something like 36 deg phase), but maybe similar to CHARD Y the optical gain is increasing and causing the loop to go unstable. There looks to be 6 dB of gain margin from the zero phase crossing.
• The DC NSUM counts have increased about 30% since the start of the lock. Assuming the sideband follows the carrier, it could follow that 30% optical gain increase would destabilize the loop
• and while I was pondering this, CHARD Y rang up very fast at 1.3 Hz and we lost lock