[evlatests] Panchromatic EVLA -- a remarkable result!

[Rick Perley]

    In preparation for an 'all bands' flux density-type run (from which 
many useful system parameters can be extracted), I ran a test which 
attempted to tune the EVLA to 26 different frequencies (in 13 pairs), in 
all 8 bands.  Each observation was for a single minute only.  Referenced 
pointing (for 3 minutes) was applied, with no 'dummy' scans.  OSRO1 mode 
was used throughout. 
    Each observation for the 13 different tuning setups was for 1 
minute.  The frequencies selected for the test were as follows.  The 
observations were made in this same order:

L-Band
    1010 and 1275 MHz (AC, BD)
    1465 and 1750
S-Band
    2050 and 2565
    2900 and 3500
C-Band
    4050 and 4885
    5700 and 6750
X-Band
    Ref. Pnt at 8100 and 8700 MHz
    8050 and 8435
    9465 and 11210
Ku-Band
    13000 and 14965
K-Band
    18585 and 22460
    24000 and 26100
Ka-Band
    29735 and 36435
Q-Band
    43340 and 48365

    The goals were:

    1) To determine if the array can indeed tune to all these frequencies
    2) To determine how long it takes to change from band to band, and 
frequency to frequency within a band
    3) To uncover any other limitations which reduce on-source integration
    4) To find which antennas do not operate at a particular band (when 
it should), or do not tune, when it should.

    Results:

    Overall, the experiment was a spectacular success.  All frequencies 
tuned successfully on nearly all antennas which are equipped with the 
necessary receivers.  The time to change frequency is no longer -- and 
usually shorter -- than the time taken by VLA antennas for similar 
operations.   The major loss of time is that associated with changing 
subband -- the last 18 seconds of some scans at some IFs were lost. 

    In more detail:

    a)  1010 MHz.  I used a 32 MHz-wide subband filter, to ensure the 
DME and radar signals were excluded.  All antennas fringed -- seven of 
the eight antennas equipped with the wide-band OMTs fringed much more 
strongly.  The exception is antenna 7, whose fringe power is 1/4 of what 
it should be.  Something is not right with this antenna. 
      A full 50 seconds on good data were obtained.  The second subband, 
at 1270 MHz fringed well on all antennas which have L-band receivers.

    b) 1465/1750 MHz:  Antennas 4 and 27 did not fringe at 1465 MHz.  
Antenna 6 failed to fringe at 1750 at one of the two scans taken.  
Otherwise, 50 seconds of good data were recorded.

    c) 2050/2565 MHz:  Antenna 22 did not fringe (apparently, the 
receiver is back for repairs).  All other antennas equipped with Sband 
receiveers fringed well.  40 seconds of good data were obtained.  
Antenna 24 failed to fringe at 2050 Mhz, on one of the two scans. 

    d) 2900/3500 MHz:  Only antenna 22 failed to fringe (of those 
equipped).  55 seconds of good data. 

    e) 4050/4885 MHz.  Antenna 13 has no receiver, otherwise, all 
antennas fringed.  Antenna 24 failed to fringe on one of two scans at 
4050 MHz.  The known RFI at 4050 MHz caused no apparent degradation (we 
were looking to the north).  50 seconds of good data. 

    f) 5700/6750 MHz.  All good data from all antennas, except 13.  50 
seconds of good data. 

    g) 8050/8435 MHz.  Antenna 6 failed to provide fringes at the upper 
frequency (BD).  55 seconds of good data. 

    h) 9465/11210 MHz:  Only 7 and 14 are equipped with the new wideband 
X-band receivers -- these fringed by far the strongest of all.  At 9465, 
all antennas fringed (but only 7 and 14 are strong).  At 11210 MHz, only 
four antennas gave fringes at all:  7, 12, 13 and 14.   The last 18 
seconds of both IFs are bad -- due apparently to the following band 
being Ku.  I thus got 32 seconds (50 - 18) of good data. 

    i) 13000/14965 MHz:  Antennas 10, 12 20 22 and 24 gave good 
fringes.   A curious asymmetry was seen:  The last 16 seconds in IF 1 
(13000 MHz) were good, but were bad in IF 2 (14965).  The following band 
was K band.  I got 40 seconds good data in IF1, and 23 seconds in IF 2.

    j) 18585/22460 MHz:  No fringes from 6, 11, 13 or 16 in IF AC 
(18585).  In IF 2, only antenna 13 failed to fringe.  Another curious 
asymmetry:  The last 18 seconds of IF 1 was bad, but was good in IF 2!  
(The following band was also K-band, but at 24000 and 26100 MHz).   
Thus, 28 seconds good data in IF 1, and 45 seconds in IF 2.
    k) 29735/36435 MHz:  No fringes from antennas 6, 13 or 16 in IF 1, 
and 13 and 16 in IF 2.  48 seconds of good data from both IFs. 

    l)  43340 and 48365 MHz:  No fringes from antennas 3, 6, 13, 16 or 
17 in IF 1, nor from antennas 3, 13 and 17 in IF 2.  Antenna 25 was weak 
at both IFs, antenna 10 was weak in IF 2.  (Probably a pointing issue).  
The last 18 seconds was bad in IF 1 only -- IF 2 fringed smartly through 
to the end.  The following band was L-band.  So, 30 seconds good data in 
IF 1, and 48 seconds in IF 2. 
   
    The mystery to me is the curious pattern of 'last 18 seconds' of bad 
data on some IFs, but not on others.  To summarize:

    X-band:    both IFs bad for last 18 seconds.  Following band was Ku.
    Ku-band:  IF 1 o.k. to the end, IF 2 bad for last 16 seconds.  
Following band was K.
    K-band:  For both tunings within this band, the last 18 seconds of 
IF 1 only were bad. 
    Q-band:  The last 18 seconds of IF 1 only were bad.  The following 
band was L. 

    All other bands gave good fringes on both IFs to the end of each scan.