[evlatests] Tests, Feb 10 & 11

[Vivek Dhawan]

Summary: In contrast to Feb 7th, where things fell apart over time,
things were quirky on the 10th, but improved on the 11th and looked
pretty good by the end.

Unfortunately, I don't know whodunit or how, on either occasion.

Read on for the blood and gore - Vivek.

------------------------------------------------------------------

1. Ken S. reported (and confirmed from his notes) that ant 16
    L band, IFs A,C were not fringing around 2PM, Feb 10th.

2. Project AB1199, ran just after that, from 3PM-midnight.
    L band was in use, line mode. I checked the calibrator, and found
    ant 16 good throughout, all 4 IFs. But 14 B,D had come unfringed.

    Auto-correlation bandpasses looked the same on all antennas/IFs.
    (coarse 7-channel spectra, so not very useful)

    One guess is that the L302's are unlocking. Ken saw some of this,
    but details of the monitor log have not been examined. Is it worth
    doing?

3. Tests of Feb 11, 4-7AM: I did the same as on Feb 7th, cycling
    over L,C,X,K bands, line and continuum.

Continuum:

    X   13, 14, 16, IFs A,B,C,D good.

    C   same

    K   14, 16, A,B,C,D: good on 2 scans, fringeless in between.

    L   same as X - fringes all around.

        BUT, there is a new quirk: Looks like closure error - oscillating
        amplitudes on the triad 13,14,16, IF A only. Cause yet unknown.

        RFI maybe?
        I used 1415MHz continuum here. In the line setup (see next item)
        I used 1435, saw no spurs and saw no closure error either.

Line:

    All antennas, all IFs, all bands produced fringes. There were a few
    small oddities, noted just in case:

    16 L: had a dropout for ~1min in a 5-min scan, recovered.

    14, 16 X: had a fringeless scan between 2 good ones.

    Closure errors seen in the continuum L band IF A were Not present.


4. Power supply spurs at ~700KHz were just barely visible in autocorr
    spectra, about 1-2% of the noise (at the backend), 1-2 channels at
    100KHz resolution. Much weaker today than on the 7th.
    Not seen in cross-power, 10-15min integration.

    The evidence is weak, because the spurs are weak, but they seem to be
    band and IF dependent. Is this just because of differing amounts of
    RF power covering them up?

    If Mike Revnell wants details, they are:

    X band: 14 A,C. 16 D.

    C band: 14 A,C,D

    K band: 16 A,C.

    L band: 14 A,C

------------------------------------------------------------------------