[Tanaka, Chen, Sugioka, Michimura, Komori]

Abstract:

We successfully recovered the DC readout state with an input power of approximately 5 W and maintained it for more than one hour.

Details:

One of the main issues preventing stable DC readout operation at high input power is an oscillation around 0.6 Hz, as reported in klog:36216.
Even after tuning the L2P drive alignment gain (klog:36218), we were unable to maintain the RF-locked state for an extended period with input power above 6 W.
Therefore, we focused on achieving DC readout operation with an input power of 5 W, where the 0.6 Hz oscillation does not occur.

We successfully maintained the DC readout state for more than one hour after performing the following steps:

• CARM gain tuning
  We tuned the CARM gain by frequently monitoring the CARM openloop transfer function while adjusting the input power using the HWP at the PSL and the reflected power on the REFL PD using the HWP at the IFO REFL port.
  We set the target angles of the PSL and REFL HWPs to 159° and 157°, respectively, resulting in the REFL PD power of approximately 3 mW and the CARM loop UGF of about 30 kHz.

• PRCL and MICH gain tuning
  We measured the openloop transfer functions of PRCL and MICH in the RF-locked state with 5 W input power.
  Since both gains were already comparable to the previous configuration without FM8 (−2.5 dB at 10 W) in PRCL1 and MICH1 (Figs. 1 and 2), we did not engage FM8 and updated the corresponding guardian settings.

• DARM offset tuning
  We changed the DARM offset in the ENGAGE_DARM_OFFSET state to −3e-5, resulting in an OMC transmission power of approximately 15 mW.

After these adjustments, we reached the DC readout state and performed the following actions:

• DARM gain tuning
  Initial measurements showed that the DARM openloop gain was 5–6 dB higher than before (the brown dots in Fig. 3).
  Therefore, we stopped engaging FM5 (+3.123 dB) in LSC-OMC_DC and reduced the overall gain from 1 to 0.8, recovering a DARM UGF of approximately 100 Hz (Fig. 3).
  This indicates that the DARM optical gain or actuator gain has increased significantly, likely due to reduced arm loss in room-temperature operation.
  The PRCL and MICH gains in DC readout remained similar to previous values without additional tuning (Figs. 4 and 5).

• BPC offset adjustment
  The initial DC readout lock was lost due to drift of the beam spot on EY in a yaw direction.
  We removed the EY yaw BPC offset of 0.09, after which the DC readout became stable.
  HThe EY and IY yaw BPC error signals at the start of DC readout remain at 0.1–0.2, which is not yet understood.

• ASC offset tuning
  We observed that fluctuations in the arm transmission power were linearly correlated with the EX TM yaw oplev signal and that the AS camera image was split in the yaw direction.
  Based on this observation, we set the DHARD yaw offset to 0.01 and these issues were solved.

Finally, we note that the DARM sensitivity in the 6–20 Hz band improved by a factor of 3–10.
Although the expected change in actuator efficiency between cryogenic and room-temperature operation is on the order of 10%, the sensitivity in this frequency range appears to have genuinely improved.
This should be confirmed, and the underlying cause should be investigated further.