[evlatests] Referenced Pointing Tests

[Rick Perley]

    Ken and I also tested referenced pointing, following the 'dropout' 
tests previous reported.

    For the pointing tests, we observed a source at elevation 72 
degrees.  Tests were done
at X and K bands. 

    A) X-band. 

    The intention was to:
       a) Observe the source with no pointing correction
       b) Determine the pointing offsets
       c) Re-observe, with offsets applied
       d) Re-determine the pointing offsets, with the offsets applied 
(this should give us no further corrections)
       e) Re-observe, with updated offsets applied.

    Due to a blunder on my part, (d) did not apply the previously 
determined offsets, so we expected to get
the same results as in (b).

    The results of this sequence were:

    Offsets were found for four EVLA antennas:  13, 14, 16 and 18.  
(Antenna 24 was locked out, and not
available).  Tests (b) and (d) gave the same results (as expected).  The 
largest offset was 0.5 arcminutes, on
antenna 16 (azimuth), for which an amplitude correction of less than 1% 
results.  Calculation of the antenna
gains, before and after the corrections are applied, show a marginal 
(0.5%) increase in amplitude, about
at the noise level. 

    B)  K-Band. 

    The sequence here was:

    a) Observe the source with no pointing corrections applied.
    b) Observe the source with the pointing corrections determined above 
(from X-band) applied.
    c) Determine 'secondary' pointing corrections (with letter code 'S' 
in the column following the bandwidth codes)
    d) Observe the source with the 'secondary' pointing correction 
applied. 
    e) Determine primary pointing corrections, at K-band, with no prior 
corrections applied.
    f) Observe the source with these K-band primary determinations 
applied. 

    Results:

    Antennas 14 and 16 only gave solutions.  No K-band fringes were seen 
from 13 or 18.  Also, 14B gave no
fringes. 

    Applying the X-band pointing corrections (about 0.5 arcminutes for 
both 14 and 16) caused a significant
*decrease* in antenna 14's amplitudes (by about 5%), but increased 
antenna 16's amplitudes by 2%. 

    The 'secondary' pointing solution returned a significant pointing 
offset in 14 (~ 0.5 arcminutes), and a much
smaller one for 16 (0.2 arcmnutes).  Application of these incremental 
solutions nicely increased antenna 14's
amplitudes to a level slightly better than the first observation (with 
no corrections applied).  No significant
change in antenna 16's amplitudes were noted. 

    The pointing solutions from the K-band alone determination ((e), 
above) returned values which are quite
different than the X-band alone, and the K-band updated solutions.  For 
antenna 14, the offset is quite small -- 0.15 arcminutes -- while for 16 
it is about 0.4 arcminutes.  Application of these K-band alone
solutions gave amplitudes which for antenna 14 are the same as with no 
corrections, and for 16 are only slightly
better than with no corrections.  This null result is expected for 14 
(where a 0.15 arcminute error results in about
a 1% amplitude loss -- too small to be securely detected), but is not 
expected for 16 (where a 0.4 arcminute error gives a 5% loss in 
amplitude -- which should be easily seen). 


    For comparison, we review the results of the VLA antennas.  In 
general, application of the X-band pointing
corrections improved VLA amplitudes (but there are a few exceptions -- 
undoubtedly due to incorrect K-X
collimation).  In *all* VLA cases, application of secondary corrections 
increased the amplitudes.  Finally,
the K-band alone solutions when applied resulted in amplitudes which are 
at the same level as those gained
from determination of secondary referenced pointing, and thus better in 
all cases than those without any
referenced pointing at all.