Some Thoughts on the LSA/MMA Antenna Diameter Issue Jeff Mangum 10/30/98 In the following I will restrict my comments to a comparison between the 10m and 12m diameter antennas under consideration. Imaging: >From the simulations and calculations that Mark Holdaway has done, there are definitely some differences between the imaging characteristics of arrays with 8m and 12m diameter elements. There have been, to my knowledge, no simulations of arrays with 10m diameter elements. From MMA Memo 178, which compared the 8m and 12m arrays, at 650 GHz image dynamic range and median fidelity is about a factor of two better, but first moment fidelity are about the same, for the 8m array. This just says that the 12m array does a better job of imaging compact sources, and the 8m array does a better job of imaging extended sources. No surprise there. These differences just reflect the relative pointing uncertainties of the two element sizes. I think that the important point here is that we *will* be able to image extended sources with the array of 12m antennas, just not quite as well as we can with an array of 8m antennas. As I will point out below, we can improve pointing by adding some innovation to the antenna design, so the imaging properties of the larger antennas can likely be made better. The Europeans have already made the case for the need for large dishes to do non-mosaic science (such as studies of high-z galaxies or protostellar cores). Looking at the compromises we need to make to be able to do both types of science, it seems to me that we give up less overall capability by going to the larger dishes. Receivers: Even though receiver technology will likely make the LSA/MMA receivers much more reliable, efficient, and less expensive, we still have the "times N" problem. With N array elements, you will have N receiver systems no matter what you do to improve the detectors and electronics. The hard part will not be producing these systems, the hard part will be maintaining these systems. Our experience at the 12 Meter Telescope is that the cryogenic systems are the most difficult to maintain. It will be *much* easier to maintain 64 systems (corresponding to the 12m element array) than it will be to maintain 90 systems (corresponding to the 10m element array). Pointing: There are many techniques that one can use to improve pointing. The antenna group has already begun to design-in several of these features. Tiltmeters, CFRP reference pointing structures, laser quadrant detectors, laser metrology systems, and thermistor systems can all be added, at minimal cost, to the antenna system to improve the pointing characteristics. For a well designed and built telescope, the telescope structure can be studied and understood at a level which can allow a detailed knowledge of the pointing behavior under a variety of conditions. I see no problems which cannot be overcome in this area. One problem area for small-dish arrays is being able to find enough pointing sources. Robert Lucas has investigated this in his LSA memo on "Reference Pointing of LSA/MMA Antennas", where he points out that the minimum usable pointing source flux is given by... S = (25/deltax)*[(15/D)**3]*[(40/N)**0.5] ...where deltax is the required pointing accuracy (usually taken as some fraction of the beam width), D is the dish diameter, and N is the number of elements. Lucas' calculations show that the 64x12m array has about 30% more usable pointing sources than the 90x10m array at 90 GHz. This difference will likely widen at the higher frequencies as the number density of sources decreases. Operation: Simply put, it is easier to maintain and operate a smaller number of array elements. Even though we will strive for uniformity, there will be differences between the array elements which limit our ability to apply global solutions to certain problems. For example, I think that uniformity in cryogenic systems will be difficult to achieve. A smoother and more efficient operation will be achievable if we have a smaller number of array elements.