Report from MMA Advisory Committee Meeting, 1993
I. Areas to watch out for.
It seemed to many of us that the original specs for the instrument, designed to please everyone at the science workshop, were in some cases pushing design efforts or planning in peculiar ways. While we do not necessarily advocate dropping these science goals, we identify here some areas where the resulting specs should not be considered essential, but instead should be evaluated in terms of their cost, both in dollars and to performance in other areas.
1. Dual frequency and dual polarization. It seems reasonable to allow operation in two channels, which might be dual polarization at a single frequency, or one polarization from each of two frequencies. The spec. should not be interpreted as two frequencies times two polarizations if this complicates design severely. For example, overall Tsys should not be sacrificed for the latter goal.
2. Observing the Sun. This should not drive receiver design in any way (there was discussion of using multiple junctions to maintain dynamic range). The committe believes that the signal-to-noise on the Sun will be so large that attenuation before the receiver would be an adequate solution. We also believe that, if observing the Sun impacts other areas of design in a major way, then the need to observe the Sun should be reevaluated.
3. The 30-50 GHz band. In current designs for optics, this band is driving the design since it requires the largest optics. The committee believes that this band should not be allowed to drive the optical design; in particular, other bands should not be compromised for this one. For example, the committee noted that this band need not be incorporated into the same optical path and dewar as the other systems and need not be one of the bands to be observed simultaneously if it proves more convenient to separate it.
4. There was some design effort to dump the image sideband to a cold load. The committee reemphasizes that both sidebands are of value and recommends that receiver design focus on sideband separation mixers which deliver both sidebands, without the noise from the other sideband.
5. Single-dish mode for all antennas. While desirable, this option should be evaluated to see how much cost/complexity could be saved by outfitting only some antennas for single-dish observing.
6. Fast-switching for phase calibration. This idea seemed to be driving the slant-axis telescope design. We believe that the need for fast switching depends on the outcome of tests of other means of phase calibration. In addition we are not convinced that the slant-axis design is clearly better. We recommend a more balanced look at the slant-axis versus more traditional designs. For example, one should first examine if the design for the SMA antennas could simply be scaled up. In addition, we note that the volume available for receivers and the simplicity of the optical path should also be considered in evaluating telescope designs.
II. Suggestions
1. The preliminary optical layout looked very complicated, partly as a result of trying to satisfy all the specs. Losses in the optics will compromise performance, and very oversize optics are needed to avoid losses. The committee recommended that a cleaner layout be attempted. There was some thought that a faster system might be easier to work with.
2. Since continuum work would benefit from wider bandwidth, the committee recommended consideration of a broadband correlator, or perhaps a set of 1GHz wide correlators, to take advantage of the very wide bandwidths available at low frequencies.
3. The committee felt that the highest priority was to assess methods of phase correction. In particular, the data from the SMA should be looked at in conjuction with the CSO tipper (it emerged from discussion that the load at the CSO tipper may not be adequate for this).
III. Answers to questions posed to the committee.
1. Phase stability measurements are necessary at least on Mauna Kea and South Baldy. Springerville looks like a backup at best and should get lowest priority. It would be very desirable to correlate phase stability with Tsys fluctuations, at least for one site, to test this method of assessing sites. In addition, the possibility of testing phase stability on a longer baseline, like 1-3 km, should be explored.
2. Further tests of phase calibration techniques should focus on the Tsys variation method and a method described by D. Woody, which uses the profile of the 22 GHz line.
3. The tests needed are those which use the BIMA facility or, if possible, the VLA. This is distinct from the site testing addressed in question 1.
4. A prototype antenna is premature at this point. The next step is to look hard at existing designs. If a scaled-up version of the SMA antenna is chosen, prototypes will soon exist. On the other hand, if a new design, like the slant-axis, proves to be superior, a prototype would probably be warranted. Note that a prototype of a completely engineered design could be useful at a later date in either case.
5. Very few compromises should be made to implement the 30-50 GHz system. In particular, it should not compromise performance at the higher frequencies, as discussed in I.3.
6. Parallel development of SIS and HFET devices is clearly appropriate, and HFETs for higher frequencies should not be ruled out.
7. See I.1. The essential point is not to compromise unduly the performance at any one band in pursuit of simultaneous observations.
8. Linear polarization for the receivers is fine.
9. Phase correction tests and testing of phase stability on sites should receive highest priority for the next 12-24 months. This includes using the BIMA setup to study Tsys-phase correlation and to see how the long baselines at BIMA work. In addition, the receiver group should look into sideband separation mixers.
10. Another scientific workshop is not needed now. We recommend one after 1-2 years, when a more definite design, with tradeoffs identified, is available and after the existing arrays have done some science with their new, expanded configurations.
Summarized by Neal Evans, 9/20/93