[evlatests] Referenced Pointing Tests

[Rick Perley]

    We did a careful check of referenced pointing today, both primary 
and secondary. For those who tire of reading these long descriptions, 
the 'executive summary' is:

    Referenced pointing works. 

    For those thirsting for details, read on ...

    Because we distrust summer weather and what it can do to high 
frequency precision observing, Ken and I elected to do these tests at C 
and X bands, with an artificially induced error of 1 arcminute in each 
of az and el, on antenna 14.  Ken did this.  (I presume he has taken in 
out, subsequently ... :) )
    The tests were done on 0555+398 -- at a very high eleveation of 78 
degrees, so the presence (or not!) of the Cos(El) term in the 
application of the pointing offset will be easily seen. 
    We observed in pointing mode at C and X bands to see what the 
pointing program found -- the large offsets were easily seen ( as it 
turns out, we have a large collimation error between X and C bands for 
that antenna, so the C-band solutions were notably different -- a red 
herring for these tests). 
    The determined solutions from X band were applied to an X-band 
observation.  The amplitudes increased precisely by the anticipated 
factor (13 percent), and a subsequent pointing confirmed there was no 
residual pointing error. 
    We then applied the same X-band solution to a C-band observation -- 
due to the large collimation error, the amplitudes were not much 
improved.  A subsequent observation at C-band confirmed that the antenna 
had moved by the amount specified by the X-band offsets. 
    A 'secondary' referenced pointing solution was then determined, and 
applied to a following C-band observation.  The amplitudes increased by 
(close to) the expected factor, and a subsequent pointing determination 
confirmed there was no residual pointing error. 
    The experiment concluded with an observation and pointing 
determination without any corrections -- to ensure that the offsets 
determined in the beginning were still the same.  They were.

    We conclude that referenced pointing now works (really!)