[evlatests] L-Band behavior, today

[Rick Perley]

    My standard stress test was run at noon today.  I'll report on L and 
P bands.  Brigette will do C and X bands.  Whoever has time will do K 
and Q. 

    At L-Band for the EVLA antennas: 

    1) Antennas 13B, and 14A and C were flagged out wholly or mostly.  
No idea why.  The other IFs were fine. 

    2) The A/C phase changed between the first and second scans (same 
source, frequency and BW), but not between the 2nd and 3rd, nor the 3rd 
and 4th.  (Four successive scans were done, to encourage bad 
behavior).   The B/D phases remained steady. 

    Combined with my polarization results, where the A/C phase changed 
up on each return to a given band, and the recent holography results, 
which show excellent stability when we stay at one band, it seems that 
the A/C phase reset is trigged by a band change, not a source change.  
The only exception to this is for the first scan of a set -- but we've 
long known that the first scan 'isn't quite right', and have long 
encouraged uses to put in a 'dummy first scan' to ensure subsequent 
stable behavior. 

    3) The usual 'short shallow drops' occurred -- for these data, only 
the B/D IFs on antennas 11, 17, 24 and 26 were doing this.  (NB  -- this 
is a very non-critical problem). 

    4) EVLA antennas are (as long reported) 10 to 40% less sensitive 
than VLA antennas at the zenith.  (The situation reverses below 30 
degrees elevation).  Two EVLA antennas are notably less sensitive then 
others:  16B, and 21A and B.  For both, the sensitivity is about a 
factor of two worse than VLA.  Probably a Tsys issue. 

    For VLA antennas:

    1) Antenna 4 is up to its usual tricks -- large short dropouts on 
all four IFs.  But as it's due for upgrade in a few days, this is 
probably not worthy of attention. 

    2) 15D is showing large amplitude increases, lasting a few tens of 
seconds at a time, then returning to normal.    The data were loaded as 
correlation coefficients, and these factor of 100 increases can't be due 
to a sudden improvement in antenna sensitivity :), nor to anything 
concerning the system temperature measurement.  Some malfunction (not 
seen last week) is at play. 

    3) Antenna 12A and C is very non-stable in phase -- remarkable, 
wandering phases of amplitude about 80 degrees on timescales of minutes. 
  Some LO is drifting about.    This one needs attention!