Looking more in detail to this line at 15.4Hz, we noticed that another line close to that, at 15.7Hz, was indeed already present since November 7 on most INJ signal (see fig. 1-2). Moreover, we noticed that yesterday, right before the INJ shift started, the power inpinging the IMC_REFL_DC photodiode decreased by itself (fig. 3). Thus, we decided to increase a bit it in order to recover the standard value. However, this action didn't solve the saturation of the SSFS.

In the following lock, we noticed that those two lines, were present almost everywhere in the INJ channel, also in the alignment channels (see fig. 4), and started to increase and saturate only once the SSFS loop was closed. 

A similar problem was already observed in the past, and is due to a non optimal tuning of the working point of the IMC AA. Therefore, we decided to perform a general check of the IMC working point, both longitudinal and angular. 

We found out that most of them were similar to their previous values with the exception of the offset for the FarField QD Galvo:

 To be noted that a similar check has already been done few week ago, when no big changes were needed. 

Anyway, once the tuning finished, both lines at ~15Hz were not present anymore (see fig. 5). 

After one unlock in locking LN3 (to be checked the reason), we could relock in LN3.