Daily Image
26-01-2010
Spot the difference
| Submitter: | Stefan Wijnholds on behalf of the EMBRACE-team |
| Description: | Since the successful solar fringe measurements (see the daily image of 9 November 2009), the EMBRACE team has been working hard to build an array of tiles modified with a 950-MHz high pass filter to make them less susceptible to RFI. During the Christmas break, these tiles were used to scan the sky at 1475 MHz observing Afristar, which is a geostationary satellite positioned at 21 degrees East. Using a clever signal processing scheme, we were able to obtain the maximum achievable SNR per tile for each direction without calibrating the array itself. Since Afristar is the single dominant source at this frequency, this reveals the average tile beam pattern of a 3x3-array of uncalibrated tiles as shown in the left panel. This image shows Afristar close to the southeastern horizon, the grating response of the system in the North and the side lobe pattern. Based on our knowledge of the arrangement of the 72 antennas on each tile and the operation of the 3-bit analog beamformer, we can make the prediction of the tile beam pattern shown in the right panel. Comparison of the two images shows that even the regular grid of small speckles can be accurately predicted. This confirms our understanding of the system and demonstrates that the deviations of the individual tiles, even when they are uncalibrated, are small, otherwise these speckles would have averaged out. The resemblance improves even further if the production tolerances of the analog beamformer components are taken into account. This measurement is an important step in the validation of the analog tile beamformer and shows that EMBRACE is ready for array level system validation. |
| Copyright: | ASTRON / EMBRACE |
| Tweet |  |