[evlatests] Failing to tune

[Rick Perley]

    I ran another L-band polarization test last night.  This one cycled 
around ten frequencies, with one minute duration for each, at 12.5 MHz 
BW.  The frequencies selected go from 1350 through 1440 MHz, separated 
by 10 MHz.   Mode 'PA' was used, so there are data only for IFpair #1. 
    The data are of superb quality, except ...  some EVLA antennas 
refuse to tune to some frequencies, or are late in actually tuning to 
the frequency requested.  But the failures are far from random:

    A)  All antennas tuned on time, every time, at 1410, 1420, 1430, and 
1440 MHz. 

    B)  At 1400 MHz, antenna 21 was late to lock up on one scan.  All 
others were fine.

    C)  Antenna 23 did not tune at all at 1350, 1360, 1370 and 1380 MHz. 

    D)  Antenna 13 missed about half the scans at 1360 MHz, and a small 
fraction at 1370 MHz. 

    E)  At 1390 MHz, 'all hell' broke loose.  Six EVLA antennas (1, 11, 
16, 19, 21, and 25) missed all or part of the scans. 

    In all cases, the bad (zero fringe amplitude) part of the scan was 
the first part.  In all cases, both IFs (A and C) failed to tune 
together, and in all cases, the data were not flagged. 

    Now, further investigation offers some tantalizing clues.   On one 
scan, three EVLA antennas 'failed to tune' at the same time.  Suspecting 
that perhaps these three were still at the previous frequency, I looked 
at the amplitudes and phases of the three baselines.  Sure enough, the 
amplitudes were high -- but the phases appeared incoherent.  However, 
the amplitudes were not as high as they should have been, so the 
tentative conclusion is that the antennas were at the correct frequency, 
but the phase rotation was set incorrectly. 

    Ken is currently perusing L302 documentation ...