Take advantage of the smoothness of most
frequency-dependent phase changes.
Expect phase changes ~f  for atmosphere and
cables, ~f -2 for the ionosphere.  Frequency-
independent phase offsets are also possible.
Lobe ambiguities need not be a driver for time
between phase calibrations (although linearity
of phase changes will be).
Care must be taken to calibrate frequency-to-
frequency amplitudes (self-cal is no help).