daniel.sigg@LIGO.ORG - posted 13:17, Friday 16 October 2026 - last comment - 16:51, Friday 16 October 2026(92332)
IMC REFL Power Adjust
Sheila Daniel
The IMC REFL power (IMC_REFL_DC) was 20.2mW when unlocked. We adjusted it to 40.8mW, a good factor of 2 higher.
In detail:
Jan10, 2025: With 60.5W input we had 16.6mW (locked)
Oct 15, 2026: With 10W input we had 0.31mW (locked); scaling this to 60W input and adding 10% for the losses in the JAC, we get ~2.1mW.
In principle we would be able to run with 8x more power on the PD, since the higher order modes in reflection of the IMC have been reduced significantly by adding JAC. However this would result in a unacceptable high power when unlocked even at 2W input. So we decided to increase the power by a factor of 2 only.
I changed the guardian so that the IMC IN1 gain is 6dB lower both for acquiring lock of the IMC and once locked. This will be adjusted again in CARM to analog, incremented up by 1dB.
First screenshot shows the OLG measurement (below the frequency scale) when the IMC guardian is in acquire (IN1 gain -18dB, fast gain 0dB). This is the situation when the guardian was trying to engage the boost, I swapped the order so that it increased the gain to nominal before engaging the boost. I hope this will help avoid the annoying difficulties we've been having getting the IMC to lock.
Final settings in LOCKED are -5dB IN1 gain, 5 dB fast gain, OLG measurement attached.
Good new/ bad news
We do seem to have helped the problem where the IMC boost coming on knocks it out of lock, so that will make relocking smoother.
The bad news is that since we increased the power on the IMC refl diode, we have had a lot of trouble with the TR_CARM steps, which had been quite reliable most of this week. I've now gone back to the table and reduced the power on IMC refl back to 20mW when the IMC is offline, and reset the IN1 gain for locking to -4 and locked to 1.
Tony has relocked the IMC now and is trying the TR_CARM offset reduction again.