[evlatests] L-Band Phase Behavior

[Rick Perley]

    I repeated the test, done 2 weeks ago, to check if our mysterious 
large phase changes, which occur upon frequency change, have left us.

    They haven't. 

    But a very curious pattern has been seen, which perhaps will 
stimulate an explanation... 

    The test cycled around 5 frequencies, with one minute total duration 
for each.  The frequencies were  1240, 1360, 1485, 1612, and 1725 MHz.  
BW = 12.5 MHz, spectral line mode '4'.  Thus, every five minutes, we 
were back at the same frequency.  The test lasted about 45 minutes -- we 
have 9 observations at each frequency. 

    The AC and BD IFs were set to the same frequency. 

    Results:

    1) Referenced to a VLA antenna, *all* EVLA antennas are suffering 
large phase changes upon changing frequency.  In other words, when we 
return to the same frequency, the phase has changed by a large number. 

    2) All EVLA antennas jump by identical amounts.  This is the 
'global' phase jump seen many times before. 

    3) The opposite polarizations (R and L) for each IF jump identically. 

    4) The jumps are identical for all five frequencies of the same IF. 

    5) ***HOWEVER*** -- the jumps for the two different IFs (--which are 
tuned to the same frequency) are different and a very curious way:

       The IF 1 jumps (A and C) have two different values:  +70 and 
-110.  (Note -- the difference is 180 degrees).
       The IF 2 jumps (B and D) have two different values:  -70 and + 
110.  (!!!)  The jumps are of the same magnitude, but of opposite signs!!!

    6) There is no pattern to whether a jump is of the + variety, or the 
- variety (which is the same as the + variety with 180 degrees added).  
For example:

       At 1725 MHz, the AC jumps were by: 
        AC:  70, 70, 70, 70, 70, -110, 70, 70, -110 degrees, while the 
BD jumps were by
        BD:  -70, -70, -70, 110, -70, 110, 110, 110, 110 degrees.
       For 1240 MHz, which followed the 1725 MHz observations by one 
minute, the jumps were:
       AC:  -110, -110, -110, 70, 70, 70, -110, 70, -110
       BD:  -70, 110, 110, -70, -70, 110, -70, 110, -70

    7) Each frequency had a different pattern, but the sizes of the 
jumps were always the same (+70 or -110 in AC, -70 or +110 in BD).

    8) No frequency *ever* came back to the correct phase.  One of the 
jumps noted above always occurred between tunings to the same frequency. 

    Perhaps a valuable prize should be offered by management for the 
first person to correctly explain this peculiar behavior ...