[evlatests] Modcomp-free tests, continuum

[Rick Perley]

    The tests yesterday were done at all bands, continuum and line, in 
the usual way. 

    Some key results from the continuum data (0.418 sec average, 50 Mhz 
BW) are reported here. 

    In general, data quality is very good. 

    All bands (including U-band) work as expected, with the few 
important exceptions noted below.  

    All problems in phase connected with the delay model (ramps and 
cups) are now completely gone. 

    A single antenna phase problem is visible:
          Antenna 26 -- all IFs, have remarkable steps in phase, which 
occur only on 10-second boundaries.  Not every multiple of 10 seconds 
has a step, but most do.  All IFs behave in an identical manner.  The 
effect is very clear at K-band, where the step size is typically 50 
degrees (the jump is not a constant value).  The effect is present, but 
much harder to discern at X-band, typically 10 degrees.  (It may scale 
with frequency, but it's not convincing). 

    A single antenna amplitude problem is visible:
          Antenna 24, IFs B and D (#2), has a remarkably rhymic 
amplitude modulation at all bands.  The signal is also weak -- about 
half in amplitude of what it should be.  I'm guessing this is a delay 
problem, and the 'rhythm' is due to the movement of the delay beam as it 
is advanced forwards (but in the wrong place).  I'll check the spectral 
line data shortly to see if this guess pans out. 

    Beyond these two specific issues, I find:

    Antenna 13, IF#2 (B and D) was very weak -- essentially dead -- at 
all frequencies. 

    Antenna 5 got progressively weaker as the frequency went up --   
pointing error?  However, referenced pointing was done, and the solution 
for 5 looks perfectly fine ...

    At K-band, all worked very well, except for the weak signal from 
antenna 5.  But at Q-band, ten antennas were very weak or dead:  
17,19,21 (all of which have no receivers), 1,5,14,18,22,26, and 28.  All 
of these last seven gave good pointing solutions, so their problems here 
are likely to have other origins than pointing.  The remaining antennas 
worked very well.