Ushiba, Tanaka

## Abstract

PRFPMI was locked with DC readout. However the peaks at 325 Hz and 332 Hz appeared in DARM sensitivity. These peaks casued maybe by the oscilation of MC loop at 150 kHz.

## What we did

• After the initial alignment, we requested the LSC_LOCK guardian to go to OBSERVATION_WITHOUT_LINES. Then, PRFPMI locked with DC readout. However, the sensitivity above 1 kHz, seems to be too bad. and the peaks at 332 Hz and 325 Hz appeared (Fig.1). So we started the investigation. Not to disturb Dan-san's work at EX, we used the Yarm lock for the investigation.
• We measured the spectra of some signals related to MC and CARM loops (Fig.2). The green lines show the spectra before switching the laser, the blue lines show the spectra just after the PRFPMI lock with DC readout. As you can see, the peaks seem to be in the blue lines. 
•  We lost the lock at once, we measured the spectra when the single-bounced beam from ITMY hit on REFL PDs. The peaks seems not to be seen in any signals.  (Fig.3)
• Then, we locked only PMC and measured the spectra of the PMC feedback signal. But, there are no peaks in the spectra (the blue line in fig.4). After that, we locked IMC and measured it but the situation was not changed (the green line in fig.4). Also, we locked Yarm but the situation was not changed (the red line in fig.4)
• We measured the spectra at TEST A and B port on CARM CMS with Moku:lab to see it above 8 kHz. Fig.5 shows the spectra when cut the input and engaged COMBOOSTs, and Fig.6 shows the spectra when Y-arm locked. As you can see, the peak around 148 kHz raised when Y-arm locked. this peak is close to 16384 Hz*9 + 330 Hz. So the 330 Hz peaks maybe down convereted from this 150 kHz peak by aliasing.  
• Moreover, we found that the 150 kHz peak seems to be bleasing (Movie 1). And, according to MC OLTF in klog, there are some shape like a resonance at 150 kHz. So I tried to decrease the overall gain of Mc and CARM loops to -3dB. Then, the 150 kHz peak seems to disappear and the 330 Hz peak also disappeared (the red line in fig.2). Therefore, one of posiibilites oscillates MC loop at 150 kHz. According to the MC OLTF with the fiber laser in klog, there seems no shapes around 150 kHz. The difference of the fiber laser and the current HPL around 150 kHz is considered to the freqeuncy responce of the broadband EOM. 
• However, if we decrease the overall gain of CARM loop to -3dB, that is, UGF of MC loop decrease from 100 kHz to ~70 kHz and the UGF of CARM loop also decrease, the amount of the frequency noise suppresion decrease. Unfortunately, this is not acceptable with the current coupling from frequency noise to DARM. We need to consider how to deal with it.