[evlatests] Final choice of initial antennas for WIDAR-0

[Michael Rupen]

Hello folks --

   This represents the final version of the initial assignment of antennas
to WIDAR-0 and the PTC.  Justifications are repeated, and expanded, from the
previous e-mail, below.

   Station Rack 1: ea19  (W04)
                   ea05e (E12)
                   ea02e (E20; KaC)
                   ea09e (E24; KaC)

   Station Rack 2: ea28e (N08; KaSC)
                   ea25  (N24; Ka)
                   ea24e (W32; KaC) [Keith Morris says IFs A & C are now
                                     ready for 3-bit samplers]
                   ea01  (W16; Ka)  when we're ready for  8 antennas

   Station Rack 3: ea18  (N16)      when the PTC is shut down
                   ea23  (E36)      when the PTC is shut down
                   ea04  (W28; Ka)  when the PTC is shut down
                   ea26  (N20)      when the PTC is shut down

    appended e ==> ready for 3-bit samplers
            Ka ==> equipped with Ka band receiver
            S  ==> equipped with S  band receiver
            L  ==> equipped with "EVLA compliant" L-band receiver (OMT)
            C  ==> equipped with "EVLA compliant" C-band receiver (OMT)

So with 7 antennas we'll have:
      4 ready for 3-bit samplers
      5 with Ka-band receivers
      4 with EVLA-compliant C-band receivers
and a range of baselines on each arm.  I didn't worry about L-band for now
since that would be yet another antenna on the east arm (ea14 at E28).


The PTC will continue running with antennas 17, 18, and 23.


In graphical form:

                         N28   ea13
                       6 N24   ea25; Ka
                      12 N20   ea26
                       9 N16   ea18

                       5 N08   ea28e; KaSC

               1 W04 ea19      E04 ea11
                MP  ea15e; C
               W12 ea17        2 E12 ea05e
            8 W16 ea01; Ka        E16 ea03e; KaC
             W20 ea16            3 E20 ea02e; KaC
                                  4 E24 ea09e; KaC
        11 W28 ea04; Ka              E28 ea14; L
        7 W32 ea24; KaC               E32 ea21; C
                                    10 E36 ea23

    preceding number --> WIDAR-0 antenna
            1- 4 are in Station Rack 1
            5- 8 are in Station Rack 2
            9-11 are in Station Rack 3


Constraints considered included:

* We will initially have 7 antennas hooked up, in single pol'n mode.
   - We may be able to hook up another antenna next week, for a total of 8.
   - Further antennas wait on refurbishing Station Boards from the PTC.
     That will take a couple weeks after we dismantle the PTC, so the end of
     March *at the earliest*.

* We cannot have the same antennas hooked up to both WIDAR-0 and the PTC.
   This only matters until we turn off the PTC, i.e., when WIDAR-0 starts
   working.

* Four of the initial WIDAR-0 antennas should be ready for 3-bit samplers.
   That means four of antennas {2, 3, 5, 9, 28, 15}.

* When we switch to full pol'n mode (i.e., both RR & LL from each antenna)
   using the 8-bit samplers, we'll have half as many antennas (rounded up --
   for these purposes, half of 7 is 4, and half of 9 is 5). We will also
   have to move some fibers around. Switching fibers within a rack
   is easy, while switching fibers *between* racks is much more painful.
   So, we'd like our five-antenna subset to consist of:
       two of the four antennas normally connected to the first Station Rack
       two of the four antennas normally connected to the second Station Rack
       one of the two  antennas normally connected to the third Station Rack
   Further, we MUST populate all of the first rack before moving on to the
   second, and so forth.  So with 7 antennas we'll use two of the four in
   Rack 1 and two of the three in Rack 2, for dual pol'n work.

* We would like a wide range of baselines, both for WIDAR-0 testing
   (a range of delays and rates) and for imaging purposes.

* We would like to include a number of the new receiver bands, to allow
   testing their wideband performance.

* While we may change this antenna allocation occasionally, we should
   not do so very often.  According to Doug and Kerry, "not very often"
   means on timescales of weeks rather than days.


Many thanks for your input --

                 -- Michael