[evlatests] K-Band stability, etc.

[Rick Perley]

    Michael sent me the K-band data that he showed last week, for my 
scrutiny.  This is what I found ...

    The run is about 15 minutes on 3C84, comprising 32 individual scans 
of 30 seconds each. 

    All antennas in the database fringed except 7C -- the cause of which 
I believe has been found and repaired. 

    Curiously, antenna 19 was slow to get back on source following a 
reference pointing scan -- an extra 27 seconds. 

    Antennas 23 and 21 have rather large delay errors -- 10 to 25 nsec.  
The rest are all < 10nsec.

    Bandpass stability is really quite amazing, with pk-pk changes in 
the bandpass, for *most* antennas, being less than 0.4% from the 
average.  But there are notable exceptions:

    *  Antenna 2 shows slow large scale (tens of MHz) changes of ~1% in LCP.
    * Antenna 6 shows a slow large scale variation, of ~3% in IF B.
    * Antenna 11 has 3% upturn or downturns in the edge channels of IF B 
(within 5 MHz of the edges). 
    * Antenna 25 shows the same effect as antenna 11 -- except that the 
up- or down- turns track in time, whereas for antenna 11, different 
scans were quite different in the edge channel amplitudes.
    A number of antennas show slow large scale changes, but (except for 
those noted above), the amplitude of the changes is less than 1%. 

    Amplitude stability is 'pretty good'.  In all cases, changes in 
amplitude occur equally in all four IFs.  There are lots of 
'micro-phenomena', nearly always at the 1 to 3% level.  Most are likely 
changes in gain or pointing -- without the switched cals there is no way 
to reliably identify the origin. 
    Some gain changes are certainly attributable to the antenna beam 
wandering away from the source:

       *  Antenna 13 for the first 5 minutes -- amplitude changes by 6%.
       *  Antenna 20:  Drops by 5 percent over a couple of minutes -- 
amplitudes were restored following a subsequent reference pointing.
       *  Antenna 22:  Same as antenna 5.
       *  Antenna 28 shows very strong loss of gain over the three 5 
minute observation (between referenced pointing scans) -- gain loss of 
20% (amplitude) in one case.  Was the pointing especially bad at this time?
       *  Antenna 5 showed two or three scans with very different 
amplitudes. 
      

    Phase stability was generally pretty good, but some odd changes were 
found, where the phase of an antenna 'jumped' by typically 10 degrees.  
In all cases, these jumps occurred on scan boundaries.  Some antennas 
were particularly bad -- antenna 19 is quite exceptional (and not in a 
good way ...), showing 10 degree jumps, up and down, at every single 
scan boundary.  For most antennas, the jumps occurred only once or twice 
over the 15 minutes.