[evlatests] Amazing Phase Behavior

[Rick Perley]

    We took more data last night, to try elucidate the nature of the
source-dependent phase changes seen on antennas 14 and 16
at X-band over the weekend. 

    Speculation was that this is due to some bizarre error in
Bz.  I can absolutely rule that explanation out. 

    The data last night were taken at C-band.  Due to difficulties
with Ken's computer, not all IFs had proper power levels set,
with the result that 16C, 14A, and 14D were dead.  Useable
data were gathered from the others.

    As before, the data were taken in 50 MHz continuum, with
400 msec averaging. 

    The phases exhibited the same gross behavior as the weekend
data -- the phases of 14 and 16 jumped to a new value with every
source change.   A quick plot indicated that both antennas had
three different phase states.  AS there were 7 sources, observed
in a cyclic way, the phase states are not connected to the source
position.

    But, more careful analysis shows that each antenna-IF has
FOUR phase states -- two of them are the same (so that every
fourth source change resulted in no phase change). 

    But it gets more bizarre than that!  Read on ...

    For the first hour of this run, the phases changed, upon every
source change, by increments of (degrees) (numbers don't add up
to 360, due to rounding and atmospheric changes): (xxx = no data)

    Antenna      A                       B                     C         
            D
---------------------------------------------------------------------------------------------
    14         65,0,140,155 |   65,0,145,155 |  xxxxxxxxxx   |  155,0,90,110
    16         xxxxxxxxxx  | 155,20,55,155 | 110,10,80,160 | xxxxxxxxxxxx
------------------------------------------------------------------------------------------------

    ... but halfway through the run (1 hour), the phase of the 2nd state 
(2nd number
in each group above), suddenly changed, so the cyclic increments became:

    Antenna      A                       B                     C         
            D
---------------------------------------------------------------------------------------------
    14         0,55,140,155 |   0,55,145,155 |  xxxxxxxxxx   |  0,150,90,110
    16         xxxxxxxxxx  | 10,150,55,155 | 0,100,80,160 | xxxxxxxxxxxx
------------------------------------------------------------------------------------------------
   
    Thus, states 1, 3, and 4 repeated phase perfectly throughout.  But 
state 2
'flipped' from one level to another. 

    Note that IFs 14A and 14B are identical -- bespeaking a common 
origin.  Because
14C and 16D were dead, it was not possible to confirm this observation. 

    As before, phase behavior within any scan is good. 

    I will write later on the sensitivity, and amplitude stability. 

    Rick