After NLN was recovered after maintenance, H1:PSL-ISS_SECONDLOOP_GAIN was 3 (instead of 5 that was in observing.snap). This is set by IMC_LOCK guardian but the actual number is hard coded in lscparams.py. We looked at the number and it seems as if 3 is the right number.
input_power = {'PRXY': 2,
'MICH': 10,
'SRXY': 10,
'INCREASE_POWER': 20,
'NLN':37}
# normally MICH=10, SRXY=10, but lowering for time when HAM6 in air### ISS Final Gain Value
ISS_acquisition_gain = 0
if input_power['NLN'] > 35:
ISS_FinalGain = 3 #for 40 W
elif input_power['NLN'] > 30:
ISS_FinalGain = 5 #for 35W alog 47607
else:
ISS_FinalGain = 7 #for 30W
The nominal power of 37 W and nominal gain of 3 in lscparams.py listed above were set on June 5th according to svn log, so it's been a while. You'd think we've been running with 37W and ISS_FinalGain=3 since then, but actually today was the first time the gain was set to 3.
TJ found that the last time IMC_LOCK was reloaded was June 4th. IMC_LOCK was never reloaded after that until today, and that's the reason why we suddenly found this. This is sneaky because IMC_LOCK itself was not changed, it's just the parameter file that is read by IMC_LOCK.
Anyway, the new gain was accepted together with H1:PSL-ISS_SECONDLOOP_REFERENCE_IN_MTRX_1_4. If somebody finds that ISS 2nd loop gain is too low, change lscparams.py and reload IMC_LOCK guardian. But don't change that gain during NLN, it will break lock.