[evlatests] Early Ka-band holography (sort of)
Rick Perley
rperley at nrao.edu
Tue Jun 17 20:39:59 EDT 2008
Ken and I attempted some holography with antenna 4, using its
Ka-band receiver. To do this, Ken tricked the system into thinking the
signal from the Ka band receiver was actually from K-band. We ran this
at 26.025 GHz.
The motivation for this was a wild idea of mine, that the apparently
10K excess spillover might be due to a significant alignment error of
the feed, such that the centroid of the illumination is in fact located
partway to the edge, thus increasing the edge illumination sufficiently
to receive a significant extra contribution from the ground. The error
would have to be in elevation, such that the effective centroid of
illumination is on the down side of the antenna, so that when we tip
downwards, the spillover stays on the ground. If the offset were to the
'up' side, the excess ground noise would quickly disappear as we tip
downwards -- this is not seen. If the offset were to the side,
something in between would result.
The expected consequences of this offset would be higher sidelobes,
and a phase gradient across the primary beam.
The holography raster was done with azimuth changing fastest, with
phase calibration every two scans -- once every 6 minutes.
The holography worked well -- in amplitude. Antenna 4 fringed
nicely at 26 GHz, with equal correlation coefficient to that obtained
from the other antennas. The horizontal beam cuts show nice symmetry,
with the expected sidelobe size. The phase gradients in each horizontal
cut are flat. All is well.
The anticipated action is in the vertical direction -- and here we
have two problems:
1) It is abundantly clear that the subreflector rotation algorithm,
used to offset the 2nd order focus shift, is not turned on. There is a
strong coma lobe from the focus drift, which is dominating any other
effect giving rise to higher sidelobes in the vertical direction. This
problem can only be repaired by proper subreflector rotation.
2) To reconstruct the phase from the holography data in the
vertical, we have to select the appropriate points in the sequential
horizontal rasters. The phase stability was poor (afternoon, big
clouds, etc.), and there is some evidence that antenna 4's phase is
changing from scan to scan, more than seems reasonable from atmospheric
instability. (However, 4 is at the end of the arm, so perhaps this is
not surprising). This problem can be avoided by performing vertical
cuts, which we shall try tomorrow.
Conclusion: All looks well in the horizontal cuts, but no judgement
of illumination, or beam symmetry can yet be made in the vertical.
More information about the evlatests
mailing list