[evlatests] Summary of characteristics at X and C bands from Cyg A observations

[Rick Perley]

    I summarize here the basic characteristics at C and X bands from the 
10 hour run taken a week ago last Friday.  To recap, the observations 
were made in OSRO1 mode (2 subband pairs, full polarization), of three 
sources (3C286, Cygnus A, and a nearby calibrator), spending 3.5 minutes 
on each source, cycling through the bands, arranged in such a way that 
no two successive observations were made the same band. 

    In general (and in contrast to negative impressions that might have 
been gained from my earlier messages), the data quality is very good for 
*most* antennas. 

    Gain variations are generally small -- a few percent in amplitude at 
most -- over the 10 hours.  Such variations that occur are of two types 
-- a slow change, looking very systemic, and one-of types, where a 
single scan had a gain a few percent away from the adjacent ones. 

    Similarly, most phase changes were small, and compatible with 
atmosphere. 

    Bandpass shapes were not nearly as stable as hoped -- this is very 
likely at least partially due to the proximity of Cygnus A to the local 
calibrator.  However, in some cases, significant shape changes (~1% 
across the 128 MHz) in amplitude were seen when large, 'one-of' gain 
changes were seen. 
   
    Failure-to-tune affected ~0.4% of the X-band observations, and about 
twice as many of the C-band observations.  (An 'observation' is here 
defined as one antenna, one IF pair, at one time.   In these two 
databases, there are 53 scans, 26 antennas, and two IF pairs -- making 
53 * 26 * 2 = 2756 observations').  From an observational standpoint, a 
'failure-to-tune' means there was only noise on the baselines to the 
affected antenna/IF pair, at the affected time.  This failure *always* 
affects both polarizations on a single IF, for the entire length of a 
scan. 

    Below, I record only the 'notable events' -- amplitude or phase 
effects much larger than the normal variance. 

    A:  C-Band amplitudes

    1) A remarkable coincidence is discovered:  For those 
scans/antennas/IFpair where there was a failure to tune, a significant 
loss of amplitude, but not change in phase, is seen in the other -- 
unaffected -- IF!  I give a log below:

    ea02:  BD failed to tune once, and AC amplitudes rose by 5%
    ea03:  AC failed to tune once, and BD amplitudes rose by 7%
    ea04:  AC failed to tune once, and BD amplitudes rose by 10%
    ea09:  AC failed to tune once, and BD amplitudes rose by 8%
    ea22:  AC failed to tune once, and BD amplitudes rose by 8%
    ea25:  BD failed to tune once, and AC amplitudes dropped by 10%:  
This event is matched by a phase event -- see below
     
    2) ea 12C showed two large gain changes (10% up then 30% down), with 
no effect seen in any other IF. 

    3) ea17A showed three large gain changes (10 -- 30% in amplitude), 
17C showed one, which was coincident with one of the three changes in 17A. 

    4) 27B had its fringes decline by factors to 2 to 3 (in amplitude) 
after the first two scans.  Thereafter, it remained stable, but weak. 

    B:  C-Band Phases

       ea04, on the BD pair, showed two scans with 150 degree phase offset.
       ea08, on AC pair, showed 7 scans with highly different phases (> 
100 degrees from apparent stable phase).
       ea10, on AC pair, has no phase stable state -- each scan appears 
to be a different phase.  The BD side, shows three scans with different 
phase.
       ea11, on BD pair, also has no stable phase state.  All scans are 
different.
       ea14, on BD pair, has 5 observations with very different phase.
       ea18, is extraordary:  All four IFs are unstable, and all four 
are *identical* in their offsets. 
       ea23, on AC pair, shows three scans with wrong phase.
       ea25, on AC pair, has one scan with phase wrong by 160 degrees -- 
the same scan as noted above with the amplitudes high by 10%. 
       ea27, on all four IFs, show a single scan with a 50 degree offset. 

    C:  X-Band Amplitudes

       ea01, IF 'B' is notable less gain stable -- at a level of 3%.  
The others are much better than 1%. 
       ea05 shows two different amplitude states, differing by 10%, 
which are exactly the same in all four IFs.
       ea12, IF 'C' shows two states, the same as at C-band. 
       ea14 is much more unstable in general, at level 5 -- 10%, on all IFs.
       ea17, A and C shows gain changes with the same characteristics as 
at X-band.
       ea27, IF B, shows the same overall gain change as noted at X-band. 
      
    D:  X-Band Phases

       ea08, AC pair, is highly gain unstable.
       ea10, AC pair, ditto.  BD pair had one scan with notably 
different phase.
       ea11, BD pair, ditto.
       ea14, AC pair, four discrepant scans.  On BD pair, no stable 
phase state is identified.