[evlatests] September 29 fluke setting test

[Mark Claussen]

I looked at the data taken by Barry on Sep 29;  this test was
to look at phase jumps in different fluke settings.  I sent
out an earlier report to a few people.


The EVLA antennas in this test were only 13, 14, and 18.

I looked at the raw phases from VLA antennas 4,12,15 with the 
three EVLA antennas.

Barry's observing sequence:   8 frequency sets (see the end of the
message for frequencies / fluke settings).  Each frequency set was 
observed for 6m with 30s scans, except freq set 1.  Freq set 1 starts 
with a 5 m scan, then 6m with 30s scans.  Then after going through the 
8 frequency sets, going back to freq set 1 there is about 40 min of 
30s scans, followed by a 6m scan, followed by 6m of 30s scans.  Then 
the next 7 frequency sets (6m of 30s scans).  Then this is repeated 
twice, and the end is about 40m of 30s scans on freq set 1.  The integration
time is 3.3 seconds.

The main question we wanted to look at was did the phase jump within 
the 6 minute frequency sets (with 30s scans).  The short answer is yes.  
But this occurred mostly only on frequency set 1.     The phase on 
VLA-EVLA baselines jumped once, for example in the first six minute 
set for frequency 1.  All EVLA-VLA baselines jumped at the same time.   
All IFs jumped at the same time.  The magnitude of the phase jumps in 
IF pairs were the same (i.e. if I plot rr-ll phase for a given IF I 
see no jump). There were no phase jumps on EVLA-EVLA baselines.  There 
appears to be a jump sometime during each recurrence of frequency set 1.  
(Of course frequency set 1 was observed longer than the others.)

Other frequency sets where I see a jump are:   frequency set   4 (once 
during the entire observing file), frequency set 7 (once during the entire
observing file).   In general all EVLA antennas and all IFs jump at the same time.   
The magnitude of the phase jump is not the same for all frequency sets 
or all times.   There are times when the phase jump is in the middle of the 
30s scan.   But more often the phase jump is from 30s scan to 30s scan.

There are some other interesting happenings.   For freq set 1, during the
first two ~40m worth of 30s scans, there were no fringes on VLA-EVLA
antennas.  When the 6m scan begins there are immediately fringes on 
VLA-EVLA baselines, and continues for the 6m worth of 30s scans.  However,
during that 40m worth of 30s scans, there *are* EVLA-EVLA fringes.  For
the last set of data on freq set 1, there are fringes on EVLA-VLA antennas,
with lots of phase jumps, including ones that occur within the 6 m sets.
On EVLA-EVLA baselines during this time, sometimes there are phase jumps
corresponding to the VLA-EVLA jumps and sometimes there are no corresponding
jumps (recall that at this time there are ~40m worth of 30s scans).

Another interesting oddity:  for baselines to antenna 13 (VLA or EVLA),
there is quite often a monotonic phase ramp within the 30s scan of about 
2-3 degrees during the scan; when it occurs it seems to continue for 
quite some time; but it is not always there.

Frequencies:  lower edge

Freq set 1:    AC:   4860.1    BD:  4810.1   =>  Flukes 100.0  200.0 
               (if L6 are 3860.1 and 3810.1)
         2:    AC:   4870.1    BD:  4820.1       Flukes 110.0  210.0
         3:          4865.1         4820.1              105.0  210.0
         4:          4865.1         4815.1              105.0  205.0
         5:          4862.1         4812.1              102.0  202.0
         6:          4861.1         4811.1              101.0  201.0
         7:          4860.6         4810.6              100.5  200.5
         8:          4860.3         4810.3              100.2  200.2


I have no conclusions to draw from this look at the data.  To me, it
appears that the phase jumps are somewhat random.   I have lots of
plot, and can make more, if people are interested.


Mark