[evlatests] State of array, Sep 05.

Vivek Dhawan vdhawan at nrao.edu
Wed Sep 5 01:56:14 EDT 2007


I did a "Rick stress test" and found some oddities
(on the EVLA, not on Rick). Continuum observations
at L,P,C,X,Q,K bands, 0.41s integration time.

FILLM complained of records out of time order.
There were missing chunks of data in the K and Q scans,
but running UVSRT to restore time-baseline order did not
fix this. Nor did re-loading FILLM with flags off.
The pattern of missing data is short bursts gone on all
baselines, during a long scan with no change to the setup.
Surviving data are good. Only the Q and K band scans
(after a reference pointing scan) showed this pattern;
L,P,C,X were unaffected by this malady. I cant even guess
if it is hardware or software.

(I used online FILLM - perhaps if I reload the data from
the archive its OK? I'll try it tomorrow.)

L: 2 scans were done (the first a dummy setup scan)
    but both survived. Phases jumped between the E/VLA
    at the scan change, with no change in setup. Cant
    tell which array jumped.

    ea14 has weird phase wander (not geometry, looks like
    an unlocked oscillator) all IF's together. OK on next
    2 scans (P, C band) and wandered again at X band;
    recovered again at Q and K.

    ea11 has a phase square wave on IF's AC only. L302
    misbehaving? Seen also at P,K bands; not at other
    bands.

P: ea24: modulated in amplitude (see K band note.)
    ea11: phase modulated, (also at L, K band)

C: ea18 has the same disease as ea11 did at L band=
    IF's AC have a square wave.

X: Good! except ea14's phase went through a turn.

K: ea11 AC phases bad, as noted above.

    ea24 modulated in amplitude with square wave of 10:1
    ratio, all IF's; phases very flat however?!
    The modulation on 24 has the same character as the
    missing data (which is missing on all antennas)
    Perhaps something is causing the data to be flagged,
    but 24 is not getting flagged and so shows the
    modulation instead of having it excised?

Q: Other than missing data chunks, pretty good data.
    no square wave modulations here.

There may be more - but I'm quitting for the day!




More information about the evlatests mailing list