[evlatests] Phase transfer - check residual phase vs. W.

[Vivek Dhawan]

I was thrashing about to locate the reason we have trouble in
the tranfer of phase from cal to target. here is a recipe:

FRING(every scan), BPASS(1 scan), Self-cal(0.1min) on calibrator.

Transfer phases to targets w. CLCAL

Plot target phase vs. W in UVPLT, this shows a linear trend of phase
of ~180 deg, corresponding to ~100psec (3cm) error for the cal-target
separation of 5deg (and roughly 1/3 that for the target 1.5deg away)

UVFIX does not remove it. (Self-cal on target does of course)

My data were processed during the week when AIPS was heaving underfoot,
so I may have done something silly.  Can Emmanuel and others see if
this is present in the recent WSC loop data? is it constant? Should
only take a few minutes to check.

Vivek.