[evlatests] 'Global' phase jumps are on the EVLA.

[Vivek Dhawan]

Phase jumps of several flavours persist in EVLA and VLA data, as
summarized here:

1. Jumps occur at random times, within scans and at scan changes;
    at both C and X band.

2. An hour without a jump is common. A day without a jump is rare.

3. Phases stay mostly in one state; the return jump is minutes,
    not hours, later. The phase slope is sometimes different in
    the 'off' state.

4. Jumps of around (not exactly) 90 and 180 are common.

5. Most of the jumps are global, i.e., all EVLA (or VLA) antennas
    jump together.  The jump on IFs A-B and B-D are the same to an
    error of a few degrees.

6. In addition, there are occasions when a few VLA antennas jump,
    but the bulk remain steady. The case of a single EVLA antenna
    jumping is rare, but has been seen.


To see if the global jumps, (item 5), are on the VLA or EVLA, I
checked a phased-VLA experiment correlated against the VLBA,
(BB209, 03Dec2006).  A phase jump did occur on the EVLA (20 VLA
+ 5 EVLA antennas, jump of ~80 deg). It is not seen (<10 deg) in
the phased array sum.  Therefore, the VLA is steady and it is the
EVLA jumping. This is guilt by inaction. Another phased-VLA
experiment will be correlated soon and I will report next week.

The converse experiment, phasing up the EVLA alone and actually
seeing the jump, is under consideration.

Vivek.