[evlatests] Effects of non-focused observing at P-band

[Rick Perley]

    The title of this is a bit of a misnomer -- we're *always* out of 
focus at P-band ...

    The recent calibrator models run gave me an opportunity to see the 
effect of observing at P-band when the subreflector was even further 
from the focus than usual.  To recap -- the (approximate) focus for the 
VLA antennas lies about 50 cm behind the subreflector in its nominal 
position.   We can move the subreflector out (back) about 10 cm from 
nominal, resulting in an out-of-focus position of about 40 cm.  This is 
done for all P-band observing, and is presumably the origin of the low 
efficiency (~ 35%) of the P-band system. 

    In the flux models run, I arranged the P-band observations to follow 
Ka-band.   This required a subreflector rotation of about 30 degrees, 
and a subreflector focus change of typically 10 cm.  The rotation takes 
about 5 to 10 seconds to complete, the focus change about 20 seconds.  
But there are significant variations between antennas -- the Ka-band 
focus position varies amongst the antennas by up to 8 cm -- which allows 
us to distinguish the amplitude and phase effects when the subreflector 
is in motion. 

    I'll show some plots at next week's Test Results meeting (yes, we're 
going to have one next week!).  Here I present the 'bottom lines':

    A) There is no discernible change in gain (either amplitude or 
phase) due to subreflector rotation.  Comment:  None was expected -- the 
rotation is small (about 30 degrees), and it is already known that the 
maximum effect of the rotation is to shift the beam pointing by about 10 
arcminutes.  This is only 1/12 of the FWHM, resulting in negligible gain 
loss. 

    B) There is an easily seen loss of gain with increasing out-of-focus 
position.  Roughly, the loss of power gain as a function of displacement 
D from the 'highest' position is:  1-15(D/lambda)^2.  So, for a 
'typical' offset of 10 cm, the loss of gain, w.r.t. the best (up) 
position, is about 15%.   This is the worst case -- some antennas appear 
to have less loss. 

    C) There is a really easily seen change in phase with subreflector 
focus change.  The observed relation is a little less than 2 degrees per 
centimeter of offset -- about what is expected:  For an on-axis ray, the 
change of phase will be 360*D/lambda, or 4 degrees per centimeter of 
defocus at 90cm wavelength.  But the great majority of the radiation 
arrives from an angle, so the effective displace is reduced.  I'd have 
expected 3 degrees/cm, but 2 sounds close enough ...  The effect of this 
on observing is very small, as it is shared by all the antennas.  The 
typical spread in focus position amongst antennas is ~ 5 cm, resulting 
in only ~10 degrees of phase differential. 

    All of this is good news for VLITE.