VLA Test/Observation Coordination Meeting B.G. Clark September 16, 2004 1. eVLA 21cm P. Napier states that the L-band feed is a difficult design problem. Because the phase center is so low, the feed sticks up above the cone quite a ways, and care must be taken so that it doesn't block other feeds. The octave frequency coverage is also difficult. Compromises had to be made. In particular, a 63 inch diameter horn was adopted instead of a theoretically superior design of 72 inches diameter, giving 8db rolloff to the edge of the subreflector, whereas the shaped reflector design presumes 11.5db. Also, the feed is placed slightly outside and below the feed circle, to minimize blocking. In theory, the effective phase center of the horn is highly frequency dependent, moving by some five meters from one end of the band to the other. In practice, this effect is not too serious. For spectral line measurements, this can be compensated by movement of the subreflector of about ten cm. Or for continuum, setting at a best mean focus results in a loss of only a few percent in the reconstructed band. R. Perley has been making single dish measurements of how the feed actually performs. He has made measurements at 1325, 1425, 1675, and 1975 MHz. The beam is measured to be quite circular, and about the nominal width (eg 29.4' at 1425). The measured best focus only varies by about 3cm across these frequencies. He calibrates the system using a (large) room temperature absorber across the feed, and thereafter monitors system gain by way of the switching noise source. The results show that the system temperature at zenith at 1425 MHz is 27.8 K (increased system noise from the old, VLA style receiver raises this at the other frequencies). Measured efficiency is 0.43, constant within the measurement errors across the band. This is significantly below what was expected. PN states that he believes the cause of the low efficiency is primarily a miscalculation of the spillover around the edge of the subreflector. Nearfield estimations suggest that only about 0.65 to 0.70 of the power hits the subreflector; previous assumptions based on farfield patterns were of order 0.83. W. Brisken states that his geometric optics ray tracing software comes up with an efficiency about as measured. Srikanth is working on a full physical optics estimation. RP measured the variation of system temperature with elevation, and finds it essentially linear in sec(z), going to 42 K at 8 degrees elevation at 1425 MHz. The linear coefficient is probably a little higher than explicable by atmospheric absorption; he guesses that perhaps only nine or ten K of the 14 K increase is due to atmosphere, the rest being due to spillover. Current VLA systems have higher efficiency (about 0.54) and comparable zenith temperatures. However, they also have severe spillover problems, so the eVLA system is more sensitive at elevations less than 60 degrees, that is, over 84% of the observable sky. PN notes that we should shortly be purchasing L band feeds; he feels that there is no other viable design that we could contemplate changing to on a short timescale, so that the choice is between this 63 inch horn design and retaining the existing VLA horn/lens system. PN states that the S and C band feeds are larger in wavelengths, corresponding to a scaling of the 72 inch feed not chosen for L band. A C band feed exists. It can be tested at the low end of its band with the existing VLA C band receiver at 4.5-5 GHz. The wide band C band receiver for eVLA is nearing completion, but is not yet ready. R. Hayward suggested that if we need to test the feed before the receiver is ready, It could be tested at the top end of its band (~8 GHz) with an existing VLA X band receiver.