[evlatests] Quick EVLA System Check

[Rick Perley]

    I took 30 minutes this morning to do a quick check of all EVLA 
antennas, at all bands. 

    I used 12.5 MHz bandwidth, correlator mode '4', giving the four 
parallel hand correlations. 

    Averaging was 3.3 seconds.  The source observed was 3C48, a standard 
calibrator. 

    Primary referenced pointing was used at C, X, and K bands.  
Secondary reference pointing (with X-band as primary) was used for Q-band. 

    Results:

    For reasons unexplained, antennas 16 and 24 gave no data at all -- 
not even with all the flags turned off in the FILLM stage.  The operator 
reports the antennas were in the array, in the appropriate state. 
    Antenna 17 was locked out, so no data was returned from it. 

    L-Band:

Antenna 26 gave no usable data on any IF. 

Antennas 13, 14, 18 and 23 gave good signals, but the *amplitudes* on 
the B and D IFs (1385 MHz) were very unsteady at times -- not the same 
for each antenna, and not the same for the two IFs for a given antenna!  
The phases were fine throughout.  The amplitudes for VLA antennas were 
fine.  I have no suggestions for this peculiar behavior. 

    C-Band. 

As only 5 antennas are outfitted at C-band (13, 14, 16, 18 and 24), and 
two of these gave no data, only three antennas fringed.  
Antennas 13, 14, and 18 gave fine, stable data on both IF pairs, both 
polarizations. 

    X-Band. 

Only the three antennas noted above as non-functional -- 16, 17, and 24 
-- gave no data. 
All others (13, 14, 18, 23 and 26) gave good, fine data with no visible 
oddities. 

    K-Band.

Completely bizarre results -- for *both* the EVLA and VLA. 
Antennas 16, 17 and 24 are out, as noted above.
All othres -- including all VLA antennas -- gave very weak fringes. 
The bandpass solutions showed IFpair 1 (IFs A and C) to have an 
effective bandwidth of ~5 MHz, centered on channel 8.  IFpair 2 (IFs B 
and D) showed normal bandwidth, but were very noisy.
The antenna gain program CALIB failed on nearly all antennas, both 
IFpairs. 

    Q-band.

The same as K-band -- only weaker.     Bandpass solutions had the same 
characteristics as at K-band.  No antenna calibration was possible -- 
insufficient SNR. 

    Something very wrong here ...