[evlatests] Ad hoc Meeting to discuss a low frequency array

[Rick Perley]

    All:

    Spurred by the recent discussions, Walter has dusted off his old 
low-frequency array idea (below).  There are some strong arguments in 
favor of discussing this further.  To do this, we are calling an ad hoc 
meeting for Monday morning, at 11 AM, in the auditorium.  Walter will 
present his concept, and we can argue the case for pursuing this further. 

    Rick

-------- Original Message --------
Subject: 	The low frequency array
Date: 	Fri, 13 Mar 2009 13:50:02 -0600 (MDT)
From: 	Walter Brisken <wbrisken at aoc.nrao.edu>
To: 	Juergen Ott <jott at aoc.nrao.edu>, Rick Perley 
<rperley at aoc.nrao.edu>, Frazer Owen <fowen at aoc.nrao.edu>, Robert Dickman 
<rdickman at aoc.nrao.edu>, Dale Frail <dfrail at aoc.nrao.edu>, Chris Carilli 
<ccarilli at aoc.nrao.edu>



Apparently there has been some interest in resurecting my thoughts on a 
low frequency array as an alternative to the other low frequency options 
that have been bandied about.  This concept was mainly meant to compete 
with the subreflector rotator modification to the VLA antennas.  In the 
past I called this the HI-z array as it would allow observations of HI 
from z=0 to ~15 barring RFI considerations.

The array:
The array would consist of twenty antennas, each a 30-m off-axis antenna 
derived by scaling up in dimensions an ATA antenna.  In the reference 
design, the feed would be a scaled ATA feed, covering 0.1 to 2 GHz.  If 
this array were designed for a more restricted frequency range a more 
efficient and/or cheaper feed for those frequencies could be used instead. 
Lopping the scaled ATA feed at the 250 GHz range would make it 
substantially smaller. The dish surface could be lightweight perforated 
panels or wire mesh.  A GMRT-style surface could be used, but may impact 
performance above 1 GHz.  The optics would be optimized for G/Tsys (unlike 
the VLA antennas which were optimized purely for gain).  This fact, 
combined with the off-axis nature of the new antennas would mean the 30-m 
dish beam would be of comparable size to the 25-m VLA antennas at the same 
frequency.  It is likely that the 20 antennas could feed into the 5 spare 
WIDAR correlator slots for use at reduced bandwidth.  This was an option 
considered for the VLBA at one point but I'm not sure if this is still 
viable.

The advantages:
1. Simultaneous observing on this array and the EVLA.
2. Observations in the 1 to 2 GHz portion of the spectrum would be able to 
use all 47 antennas simultaneously, significantly increasing the 
sensitivity at moderately redshifted HI.
3. No negative impact on VLA antennas (structural or electrical)
4. Higher gain, lower sidelobes, and cleaner beam mean a simpler imaging 
problem and substantially less sensitivity to out-of-beam RFI.  I expect
that G/Tsys will be 4 to 5 times higher than for a VLA antenna at 330 MHz.
5. Polarization properties of this antenna would be superior to the 
VLA+dipole approach.
6. This antenna would open up a much larger bandwidth than the dipole 
could.  Its improved RFI resistance may open up the 600 to 1100 MHz 
spectrum which is crucial for cosmologically redshifted sources.

Cost:
This is still rather uncertain, but extrapolating from GMRT, from the ATA 
and using somewhat standard cost scaling rules I arrive at costs somewhere 
between $10m and $20m.  This costing was done a few years ago but I don't 
have my notes on that in front of me at the moment.  It was fairly clear 
at the time that the cost of this array would be lower or at least 
competitive with the subreflector rotator.