[evlatests] antenna 24 AC

[Ken Sowinski]

Antenna 24 has not been a very good pointer.  In examining
the pointing run from last night I found that many trials
were rejected because the estimated beam width was narrower
than all other antennas by about 8 percent.  This has happened
in previous pointing runs as well.

Focus seems OK, in any case it should only make a broader beam.
However in examining focus I saw that the AC amplitudes are
not very stable.  There are changes of about 5 percent from
record to record in AC while BD is stable.  There are similar
variations in the IF A synch detector, but less clearly in IF C.
Again, both B and D synch detectors are stable.

I don't see how to consistently make a narrower beam from this
behavior, but it certainly can make the pointing unreliable.
Given the symptoms: only A and C and at X band, it just about
has to be the AC L302.  If I had to guess, it might be related
to phase switching.

If the SD variations are indeed linked to the amplitude
variations than it should be easy to see in real time whether
swapping L302s helps, even if we can't see correlated data.

Further information might be gleaned by:
1.  Using 0.4 sec integration time to see the temporal behavior.
2.  Turn off phase switching.
3.  Convince the pointing solver to only use BD data.