[evlatests] Holography Phenomena with EVLA

[Rick Perley]

We have run numerous tests of EVLA holography over the past week, as 
part of an effort to determine if optical misalignment is responsible 
for the apparently high-frequency degradation of EVLA sensitivity within 
L-band.

One of the many odd phenomena revealed in these tests is the difficulty 
in synchronization of the VLA and EVLA  antennas for holographic 
scanning.  Walter believes he has repaired this particularly annoying 
problem, and the report here is from a careful test, done this morning. 

We did two 11 x 11 rasters, with 5 points/beam.  This means we did not 
raster outside the primary beam -- the goal here was to characterize 
phase peculiarities detected in earlier tests.  Because we stayed within 
the primary beam, all phases determined should be close to zero (after 
calibration).  

The first 11 x 11 raster was taken with vertical scans.  The second was 
with horizontal scanning.   Each scan consisted of 11 equally-spaced 
pointings along the vertical, or horizontal line.  Each pointing was for 
ten seconds, but the first ~3.5 seconds cannot be used due to antenna 
motion during that interval.  Corrrelator dump time was 1.66 seconds.  
The frequencies chosen were 1275 and 1706 MHz.  We used 3C48 as the 
source, which was at an intermediate elevation (about 50 degrees) at the 
time. 

Four EVLA antennas were included -- all worked well (other than the 
peculiarities noted below).  The calibration was straightforwards, with 
no amplitude or phase problems.   Judging from the calibrator data 
alone, everything worked perfectly.   There was no significant RFI 
during this run. 

Results

A)  Amplitudes. 

All amplitude scans for EVLA antennas looked normal, with the primary 
beam of normal width and shape.  There are some occasional 'drops' in 
amplitude which affect all four EVLA antennas similarly (but no VLA 
antennas), lasting for a single integration.  I'm inclined to ignore 
these for now, as the phase issues below are considerably more 
troublesome. 

B)  Phases.

Many bizarre effects are noted:

i)  EVLA phases, within any one scan,  have an apparently arbitrary 
phase added.  The value is always the same for all EVLA antennas, for 
both polarizations, and for both frequencies.  The values are always at 
(or near) +/- 90, or 180 degrees.  There is no discernible pattern to 
these changes.  All pointings within any one scan appear to be affected 
(offset) equally.  The transition occurs between the scans.  No such 
phase offsets are noted in the calibration data (I calibrated 6 times, 
whereas it is rare for three holographic scans to have the same phase in 
succession). 

ii) There are phase changes *within* a single holographic scan.  These 
come in two types:
       a) At the lower frequency of 1275 MHz (where we are certainly 
within the main beam for the offset step size of 6.46 arcmintes), 
signficant phase offsets are commonly seen on all EVLA antennas on the 
end positions only (that is, on the first and last positions of a 
scan).  It is not possible to determine if these occur at the higher 
frequency, since the smaller primary beam makes these positions lie 
outside the primary beam.   From the data at hand, it is not impossible 
that these phases changes represent real beam phase.  I'll have to run 
the program with a smaller stepsize to confirm this.
       b) The phase can change dramatically within the middle of a 
holographic scan.  This is not common, but when it happens, it happens 
for a single pointing, both frequencies, both IFs and all EVLA antennas. 

    I'll emphasize that proper holography cannot be done with these 
phase characteristics! 

    The only other conclusion I can make is that the 'phase curvature' I 
reported at the test meeting today, wherein a parabolic phase was seen 
across the main beam at 1675 MHz, is confirmed for 1706 MHz.  However, 
it is present *only* in antenna 14 (which was the only one I looked at 
in the earlier data, reported on at the meeting).