[evlatests] Pointing Tests

[Ken Sowinski]

On Fri, 19 May 2006, Ken Sowinski wrote:

>
>
> On Fri, 19 May 2006, Rick Perley wrote:
>
>>   a) Antennas 13, 14, 16 and 18 gave a pointing solution only once -- for 
>> the first source at X-band.  All subsequent pointing offset determinations 
>> failed for these antennas (but worked for the VLA antennas).  Ken reports 
>> to me that the solutions failed because the antennas didn't `wiggle' -- 
>> they stayed on source when they should have been moving to the half power 
>> points in az and el.
>
> Rick did not mention that the second attempt was made using subarray 2

I finally realized yesterday that is is just the fact that EVLA antennas 
were in sub 2 that caused pointing to fail.  The only plausible way to 
cause the antenna not to wiggle is if waitModcomp() does not return. 
This is bound to happen if pointing antennas are in sub 2 in the Modcomp 
because waitModcomp() only inspects the on-source flag for subarray 1.
The mystery now is to explain what the first pointing scan worked.
I suppose that was because a skip was done to reconfigure the array
into two subarrays and both sub 1 and sub 2 were on source at the same 
time so waitModcomp() saw the signal from sub 1.

Fixing in this in the script would be clumsy at best because obsscript
would need an external parameter to tell it what subarray the script
was intended to use at observe time.  It is somewhat less ugly to
include a subarray when the script is launched.  This can be used to
set some member of the array object to tell the system which subarray
the Modcomps are using for the .obs equivalant of this script.

This only explains what happened Friday afternoon.  We still need to
understand why Rick claims referenced pointing is not working.