Chromatism baseline ripple

As became clear in the discussion of the stray radiation, it is possible that radiation reaches the feed through more than one path. When this happens the pathlength differences upon reaching the detector will cause interference, the intensity of which will be frequency dependent. In spectral line observation, where the total bandwidth is analysed into several hundreds of separate channels, this interference can lead to a periodic ripple, normally called baseline ripple. Weinreb (1967) was the first to draw attention to this fact. While baseline ripple can also be caused by mismatches within the receiver electronics, we limit ourselves here to shortly describe the causes of baseline ripple which are connected to the antenna and its feed. A thorough treatment of this case has been presented by Morris (1978). Most important is the backscatter-ing of radiation by the feed and its surrounding mounting plate. Some of this radiation is reflected again by the reflector and via the feed support structure to re-enter the feed, where it will give rise to the interference mentioned above. The amplitude of the ripple is proportional to the wavelength. Possible paths are illustrated in Fig. 6.14. Most serious is path A, the specular reflection from the center region of the reflector.

Fig. 6.14 - Sketch of three paths along which radiation scattered from the feed and its surroundings can be returned to the feed and cause interference with the directly received signal. (After Morris, 1978)

Depending on their structural layout, the support legs can contribute significantly to the ripple. The "periodicity" of the ripple caused by path A will be c/2 f, where c is the velocity of light and f the focal length of the antenna. Thus this will be 10 - 30 MHz for typical antennas of 10 - 50 m diameter. In the case of a Cassegrain configuration, we need to replace f by m-f with m the magnification factor. Thus with typical values of 10 < m < 20 the ripple will be in the lower MHz range. The ripple due to path A can be significantly suppressed by combining two observations taken with the feed axially defocussed at plus and minus 1/8 from the focal point. The 1/2 total path difference between the doubly reflected rays will effectively cancel the ripple effect. This is illustrated in Fig. 6.15. The multi-path reflections via the support legs will not be completely cancelled this way. Methods to reduce the chromatism are described by Padman and Hills (1991).

1, it 'V1 „■v JVW 1

0 0

Post a comment