Minutes of MNRF Phase Correction Meeting, 18 June 1999
Present: J. Brooks, M. Sinclair, G. Carrad, R. Sault, R. Manchester, L. Staveley-Smith, W. Wieringa (by phone), P. Hall (chair)
1. Bob Sault gave a summary of the phase correction material presented at the recent mm imaging conference in Tucson. The results of Dave Woody and the OVRO group are impressive, with rms path corrections of 50 um (about 5 deg at 100 GHz) being obtained with the uncooled (200 K Tsys) receiver system. Significantly, phase can be tracked between astronomical calibrators and target sources, and correction has been demonstrated under at least light overcast.
The ATNF 4-channel phase correction system is broadly similar to the OVRO design, giving some cause for optimism about our prospects for significantly improving the Narrabri mm-wave data.
Bob Sault commented that Woody believes it is still appropriate to take the IRAM approach and record both corrected and uncorrected data.
The 183 GHz remote sensing approach is not in very good shape. Saturation is a severe problem at Mauna Kea and is likely to remain so at the ALMA site. Also, the prototype JCMT-CSO system used a DSB mixer, effectively preventing effective rejection of, for example, ground pickup. A cooled 22 GHz system would most likely provide the required path length sensitivity for ALMA, but 183 GHz would be much easier to accomodate in the antenna optics.
One disadvantage of 22 GHz is the increasing use of the band for satellite communications and broadcast systems. Bob Sault has produced an optimization for filter centre frequencies, taking into account the Narrabri atmosphere (based on met data, including Moree balloon launches) and our method of using WLM data in interferometric correction. The optimizations are fairly broad however and the suggested centre frequencies may need to be shifted to avoid interference.
2. There was some discussion about phase correction possibilities with the (narrowband) prototype receivers. The feeling is that the ~2 GHz system bandwidth is too narrow to do any meaningful phase correction measurements, but it will still be useful to assess the receiver stability. Mal Sinclair expects full-band MMIC amplifiers to be available by mid 2000.
3. Graeme Carrad outlined some of the interactions he's been having with Jon Crofts of Astrowave, the Melbourne consultancy producing (by late 99) the prototype phase correction box. Graeme will include in the minutes of the next future development meeting a simplified block diagram of the system being contructed. Astrowave are presently contracted to deliver one system for bench testing, but are producing metalwork etc for two units on the assumption that interferometer trials will be undertaken in the first half of 2000. The first units will use discrete modules (amplifiers, detectors etc) but later units will probably use a hybrid planar construction style more suitable for mass production (and hopefully even more stable).
Filters will be printed substrate style devices if at all possible, giving maximum versatility in changing sounding frequencies should interference be a problem in the future.
Peter Hall noted that some of the previously-discussed issues to do with receiver cal size(s) and repetition rates remain on the books. We need to measure the receiver and WLM characteristics before we can be definitive in these areas.
There was some discussion about system cost. Graeme will include a more detailed costing in next week's future development meeting but it appears possible to implement each WLM for an all-up cost of under $25k.
4. It was noted that a survey of actual 12 mm band interference at Narrabri would be a good thing. The prototype receivers are some help in this regard, but even a simple horn/spectrum analyser combination would spot the worst of any satellite downlinks etc. P. Hall has raised this in the past with Observatory staff and will do so again.
5. Bob Sault gave a brief run down on his 22 GHz calibrator observations next week. Bob has ~30 hrs of telescope time and will look at ~ 500 sources, hopefully with up to 3 cuts per source. Most of the sources come from the PMN list; with ~ 0.5 Jy sources being common, an integration time of ~ 3 mis per source is likely. The test interferometer baseline is ~ 60 m. Mark Wieringa will assist with the observing.