Single boresight beam full-band Tsys/efficiency over 28 antennas:
Additional notes:
Strong negative drop-outs around 880 and 1650 MHz are most likely a correlator error - misalignment for a small number of channels causing loss of signal. This is being worked on by our engineers. It is not always present and does not always appear in the same place.
Small step discontinuity at 988 MHz - this is the transition between our low and mid frequency range - NOT the actual filter edge. Cause unclear.
Note that the section between 970 and 1258 MHz can be reached in both the low and mid-band filters. Data from both configuration options are plotted and are almost identical.
RFI present between roughly 1150 and 1300 MHz. Most of this is from satellite navigation signals and will be present at all times. Actual occupancy not quite as bad as it seems, could be mitigated further with flagging and interpolation during formation of beam weights. This is also being worked on.
Rapid upturn at 1530 MHz is the mid-band filter edge. Can be avoided with overlap from the high-band filter.
RFI present between 1550 and 1600 MHz, probably secondary transmission frequencies from navigation satellites.
Black traces show individual antennas and give an indication of the spread that can occur due to beamforming variations. We are developing ways to detect outlying antennas at beam formation time.
All data come from boresight beams, using observations covering 4 sections of 288 MHz each (one repeated in two different filters). There will be degradation towards the edge of the field of view and approximate scaling relations for this can be obtained from ACES memo 15.
