[evlatests] Modcomp-Free Tests, 25 May

[Rick Perley]

    Now that power is back, and all the antennas are moved to, and are 
working at, their new positions, we can restart the Modcomp-free tests. 

    A quick look was done, in the standard way.  Because the weather is 
poor, and pointing/delays are not yet accurately determined for many 
antennas, the tests were done at P, L, C, and X bands only. 

    It was immediately found that the continuum data (taken at 50 MHz) 
was not useful for testing, as the amplitudes were highly variable on 
baseline/antenna. 

    Reducing the line data quickly gave the answer -- there are very 
large delay errors amongst all antennas -- the typical value being about 
100 nsec (huge!). 

    Concentrating on the line results, I find:

    0) Antennas 14 and 22 gave no fringes at any band, or any 
frequency.  The operator confirmed that antenna 22 was out of the array, 
but states that 14 was in, and should have been working.  Antenna 19 was 
out at the time. 

    1) No sensible data from Pband at all.  Something is significantly 
wrong. 

    2) At Lband:

          - All antennas (other than 14, 19, and 22) gave good, steady 
fringes.  All bandpasses, except #25, gave normal looking bandpasses.  
Antenna 25 has an especially high delay errors-- over 250 nsec. 
          - The EVLA and VLA were again incoherent with each other, in 
IF#2 only!!!  This is a line mode effect -- there are good fringes seen 
between EVLA and VLA in the continuum data. 

    3) At CBand:

          - In addition to 14, 19 and 21, antenna 23 gave no fringes.  
No receiver, I presume.  Antenna 25's bandpass was very odd (worse than 
at L-band). 

    4) At X-Band:

       - Fringes are rather rattier than hoped -- probably weather 
related.  Bandpass from 25 is hugely weird.  (Sorry for the ad hoc term, 
but I can't think of a better descriptor).    Otherwise, all looks about 
as expected. 

    Overall:

     Stable data, on all antennas, are being provided.  The delays need 
to be straighted out before more exacting tests can be done.