We had thought and had been crossing our fingers that the 4.2 Hz problem was gone, but it just caused a lockloss from NomLowNoise.  ADS wasn't very well converged, but was most of the way there - all error signal values were at or below 0.02, although it's much nicer when they get below 0.005.

The 2 attached figures are the same data, one is zoomed in close to the lockloss time.