[Yokozawa, Komori]

Abstract:

We observe improvement of DARM sensitivity.
This is primarily due to the disappearance of several peaks in the 100–400 Hz range.
However, the sensitivity in the 60–100 Hz range remains unchanged.

Detail:

Following the successful short-term locking of the IFO using DC readout, I evaluated the DARM sensitivity (Fig.1).
The DC PD power during this measurement was 10 mW, resulting in a reduced optical gain and open-loop gain by a factor of sqrt(15/10)​ compared to the typical value.
To correctly calibrate the sensitivity, I calculated the new open-loop gain (G) and compensated for the change by multiplying by the ratio of abs(1+G).
The new DARM sensitivity reveals two main points:

- 60–100 Hz

Unfortunately, the sensitivity in this frequency range has not been improved.
Potential reasons for this are as follows:

• My estimation of the suspension thermal noise is completely wrong and some different noises are dominant. A cross-check would be greatly appreciated.
• Beam spot mis-centering: A shift in the beam spot position on the test masses could have increased the contribution of pitch thermal noise, compensating the improvements achieved by cooling to a lower temperature. The current mis-centering should be measured to confirm this.
• DAC + coil driver noise are still dominant. Only one dewhitening filter was applied to EX during this measurement. Applying two dewhitening filters to all test masses could reduce this noise source further.

- 100–400 Hz (Fig. 2):

Significant improvement in sensitivity was observed in this frequency range due to the disappearance of many peaks.
Many of these peaks were previously observed in the CARM PD signals (klog:31667) and could have contributed as CARM sensing noise or input jitter noise.
The potential reasons for this improvement include:

• Reduction in CARM sensing noise: The IFO got closer to a critically coupled cavity. The old typical REFL power of ~3 mW was reduced to 2.4 mW during this measurement.
  Further reduction to 1.2 mW has been achieved in previous work (klog:32362). It is worth trying to reduce the REFL power to this level and check the sensitivity.
• Changes in input jitter coupling: Different beam spot position could have altered the coupling of input jitter noise, leading to the observed improvements.
• Scattered light mitigation: Efforts to suppress scattered light at the PSL and REFL tables were effective.

The next steps after achieving a stable IFO operation are applying two dewhitening filters to all test masses, and optimizing the beam spot positions based on the DARM sensitivity.