We found that ISS went down freqeuntly and this lock loss often casues IFO lock loss (fig.1). The ISS lock loss was caused by the AOM saturation. According to Miyakawa-san, the saturation occured if the voltage appling AOM was more than 4 V (top panel in fig.1). This indicated the trans. power fluctuatioion became large.

And we found that IP2 PZT was saturated again. This is because I forgot to restore the PZT offset (K1:IMC-PZT2_PIT_OFFSET) which was changed in last Saturday (see in klog31940). So the lower side of the PZT range became too small, only -5 V. 

Therefore I restored the offset to the nominal value, 75 V and then I offloaded it.

After the offloading work, We locked IFO and ISS. Fig. 2 shows the timeseires related to ISS, the RMS of AOM feedback signal seems to become smaller slightly.