[evlatests] Phase clunks at K and P band and other phase behaviour

[Mark Claussen]

First time I've looked at EVLA data in a while, but I knew Vivek
was out. 

Ken took some data at K and P near the end of software time
today (March 23, 2006).  About 50 minutes at K on 3C84, and 
about 25 minutes at P on 3C48.  The 4/P converter was installed 
on An 14 yesterday, and the polarizations have been made the right 
way around this morning.  UX converter has been installed in Antenna 13.

Quick summary of phase behaviour:

K band:  

There is a phase clunk every 10s (records are 1 2/3 s)
in antenna 16, all IFs.  In addition there is a phase 
slope with time (wind) in antenna 16, all IFs, about, 
the same magnitude over 10s as the clunk... so the overall 
effect makes the phases look more or less constant (on 
short timescales; on longer timescales --- several minutes 
--- the troposphere takes over).  The phase wind and clunk 
is about 50 degrees and more or less independent of baseline 
(haven't looked very carefully at this yet).

In antenna 14, there is also a phase clunk, but only on the
BD side.  The clunk happens every 10s. The phases within one
set of 10s is reasonably constant, but the size of the
clunk here is large, and changes systematically, so that the
overall effect resembles the parabolic phase behaviour we
associate with a source position or baseline error (the systematic
phase clunk goes through several nulls --- flat parts of the
parabola in the 50 minutes of data).   Seems I remember seeing
similar behaviour some time ago.....

Antenna 13 --- no evidence of these phase clunks / winds on any IF.

P band on 14:

The AC pair phases seem pretty steady.  The BD pair has a clunk 
every 10s; within each 10s the phase monotonically changes by 
about 60 degrees or so; there is a systematic phase change from
10s chunk to 10s chunk, which appears, at this quick look to be
constant by about one turn in 1 minute.

That's probably enough for this evening.  Perhaps more later.

Mark