I've analyzed the trend of the 25MHz signals (both QD1 and QD2) during the CARM 0 lock of the last week, while the SR is closed with the DCP signal.
The trend is visible in Figure 1. The offset between the DCP and the 25MHz is not constant and depends strongly on the alignment of the other DOFs such as BS and DIFFp (at it was also expected).
Moreover there is a strange behavior occurred yesterday that has to be analyzed (see third ellipse in the plot).