Figure 1 and 2. The glitches affecting BS TY are present only on B1p QD1 50MHz and not on B1p QD2 50MHz, so it is not an interferometer problem. They are also not present on B1p QD1 56MHz, so it is not a quadrant hardware problem that should affect all demodulation frequencies. These glitches are very short, ~1ms or less. They affect both H, V and Sum channels.

Figure 3 Looking further they correspond to the phase_noise_corr jumping back and forth by 11 radians. If I am reading well where it comes from in Acl, this is due to the B1s 112MHz phi also jumping by 11 radians. I don't understand why the B1s 112MHz phase jumps.