[Ushiba, Komori]

Abstract:

We successfully removed a significant phase delay from the DSOFT Y openloop transfer function by adjusting the combination values in the input matrix of the transmission QPDs.

Details:

We had been suffering from significant coupling from other degrees of freedom in the DSOFT Y signal, as reported in klog:36325.
Today, we performed decoupling of the {D, C}SOFT modes from the DHARD mode.
The method is the same as that used in klog:36288, namely injecting a 50-mHz sine wave into DHARD Y, measuring the peak height in the {D, C}SOFT signals, and adjusting the combination of the two transmission QPDs.

Power spectra of each WFS signal before and after the adjustment are compared in Fig. 1.
As in the previous adjustment for the pitch mode, CHARD was already well decoupled from the beginning, while noticeable couplings were observed in the {D, C}SOFT modes.
Although the phase difference between QPD1 and QPD2 is far from 0 deg or 180 deg, we adjusted the input matrix values based on the difference in peak heights.

The coupling in terms of peak height could not be further improved.
However, the problematic openloop transfer function of DSOFT Y (brown in Fig. 2) was slightly improved; in other words, the phase delay was partially recovered (magenta).
After further adjustment of the input matrix values, we obtained a better openloop transfer function, as shown by the red line.
This improvement allowed a higher unity gain frequency with a larger phase margin, leading to reduced fluctuations in the EX and IX yaw motions.

We also checked the openloop transfer function of CSOFT Y and confirmed that there were no significant couplings (Fig. 3).
We slightly increased the gain of CSOFT Y by increasing the corresponding matrix values.
The input matrix values after today’s work are shown in Fig. 4.

After today’s work, the interferometer reached the observation state without any problems.