[evlatests] tests of Nov 7 & 8.

[Vivek Dhawan]

Test 1: 3 hours, a 'mini-baselines' run, X-band, restricted to
         about 20deg in elevation, 90deg in azimuth.

Test 2: Regular baseline run X-band, 5 hours.

The goals were:

a. See if the EVLA phases were better behaved over a limited range
    compared to the full baseline run with full-sky, over-the-top
    coverage.

b. See if I could get a better position for 23 (master pad) and the
    other EVLA antennas in either run.

We lost power for an hour, some antennas did not return. Test 1 gave
1-2 hours of data, Test 2 about 3hrs.

A surprise:

24 is now the best. It used to have very poor phase vs. elevation.

    Going back, I find that -
    It has not been moved since Aug.2, when it left the master pad.
    It was terrible on Sep 25 and Sep 30.
    It was GREAT on Oct 24.

    Anyone willing to fess up? And repeat the trick on the others?

Otherwise, the usual troubles:

18 is still quite good. But the position fit on Oct 24 is different
    from that on Nov 08; I don't trust it yet. (Moved Oct 17).

13, 14, 16:  bad as before - cannot fit a position.

26 worst behaved phases.

23 worked (B D) for 1.5hr, did not return from power glitch.
    Tied with 26 for worst.

Vivek.

P.S. VLA ant 10 showed no sign of problems (IF A strange phases,
      high amplitudes) reported Oct 25. Ray Ferraro reported a sampler
      replaced, seems to have fixed it. Correlator controller exonerated
      for now.