[Ushiba, Yokozawa, Komori]
Abstract:
We increased the laser power at the out-of-loop CARM PD to measure the input jitter with better signal-to-noise ratio.
Detail:
As discussed in klog:33312, we aim to subtract input jitter noise using the out-of-loop CARM signal from REFL PDA3.
To enhance the signal-to-noise ratio of the jitter signal, we increased the optical power incident on PDA3 as much as possible.
First, we rotated the half wave plate to change the power entering PDA3, and PDA4 for potential usage as REFL MICH detection.
Since the power at PDA4 was initially higher than that at PDA3, we adjusted the power at PDA4 to approximately 50 mW during PRMI 3F locking, when the power on PDs gets maximum.
This value of 50 mW is upper limit of the incident power to the PD.
Consequently, the power at PDA3 increased to roughly 35 mW—five times higher than before this adjustment.
However, we found that the interferometer could not be locked with this configuration, as significant noise was introduced during the IR handover.
Ushiba-san suspected that the cause was the well-known crosstalk between the REFL PDs and the green phase signal, which appeared to be affecting the system again in this case.
The attached figure shows a comparison of various signals with and without light incident on PDA3 and PDA4, supporting the hypothesis that crosstalk was the primary issue.
To resolve this, we completely blocked the light to PDA4—which has not been used so far—and reduced the power on PDA3 by half.
With this configuration, we successfully recovered the interferometer lock.
As a result, the optical power on PDA3 during the observation state increased from 2.8 mW before this work, to 8.2 mW.
