![[Bristol arms]](../gifs/univ_logo.gif){kind=logo}
| Updated: 8-Jun-2006 | Mark Birkinshaw |
![[Valid HTML 4.01]](/gifs/valid-html401.gif)
|
Information on the correlator can be gained from the presentation given by Derek Kubo in 2003. This shows the correlator layouts for the 6, 13, and 19-element arrays.
Some current issues with the correlator can be judged from the noise analysis of a run with ecosorb loads over the horns, done by Hiroaki Nishioka on 3 June 2006. The data are for 1L2L lags 1 to 4, data taken from 1546 (local time) for several days.
The correlator addressing system is given in the address table forwarded by Kyle on 8 June 2006.
A plot of the 1L-2L correlator lag data, with one lag output plotted against another (4 lags, 6 pairs) shows that the data describe a pattern of ellipses. It is best for the operation of the correlator if the ratio of the ellipse axes is large, since then the extracted visibilities in the two channels are almost independent. However, it can be seen that the axial ratios are not large. This should be improved by flattening the response of the receivers and flattening the input phases into the correlator.
|
| Updated: 8-Jun-2006 | Mark Birkinshaw |
|