Figure 1 shows MICH, a comb of 1Hz lines disappears when the Internal PLL is engaged around 10:30 UTC, and reappears when the External PLL lock is engaged around 16:00 UTC.
Figure 2 shows spectra just before (purple) and after (before) engaging the internal LNFS PLL. The 1Hz comb disappears from MICH, PRCL and DARM. For DARM it is less clear as other broadband noise is present, but clear the 10Hz part of the comb is gone.
That 1Hz comb is also clear on the SPRB clock and on the B4 112MHz phase, but changing the LNFS PLL doesn't impact these two.
This is due to a cabling mistake done a few weeks ago, which causes the 100 MHz and IRIG-B signal being sent to the LNFS on the same cable instead of separately.
