[evlatests] L352 meters and fiber delay measurements.

[Vivek Dhawan]

>From late February to early March, the meters were tracking the
same signal - out to antenna 13 on fiber F1, back on F2,
then split to feed the 4 meters. As reported a couple of weeks
ago, units 40 and 51 have opposite sign to 3d, otherwise these
3 units track each other within about +_1ps (one-way error).
The diurnal delay change is +_50 to +_100ps.  #52 disagreed with
the rest by about 15%, and has been returned to the AOC.

March 8-11th, #3d was connected to the signal returned on another
fiber F3 in the same bundle, while #40 and #51 continued as before.
In actual use, half of the round-trip delay will be applied to
correct the astronomical data. We are not yet ready for that,
because in this test, path F1-F2 differs by +_10ps (+_180deg phase
at 50GHz) from path F1-F3.

Vivek