| Description : |
The autoguider is keeping track of 8 bundles of 7 fibres each, hooked up to bright stars in the 2 degree field of the 2dF instrument. When the astrometry (the positions of the stars) are good, all stars should line up in the middle fibre, and the guider takes a systematic brightening on one side to mean that tracking needs to be adjusted to evenly illuminate the fibres again (you want as much light going down the middle fibre as possible, because normal fibres hooked up to normal oberved sources only come with one fibre, not 7). They usually don't line up perfectly, because of numerous factors -- not the least being that the atmosphere changes throughout the 2 hour duration of each field, distorting the view. The fibres are splayed out on the CCD detector, and although they appear separated on the detector by quite some distance, they are separated by only an arcsecond or two "on the sky", so it doesn't take big errors in the astrometry to really complicate guiding. Fortunately, we have 8 stars to improve our chances of guiding correctly. More complications are that several fibres will be broken at any one time. The guider doesn't know which fibres they are, so you have to feed hints to it so the broken fibres don't bias the tracking position. Why do I monitor the seeing (turbulence)? There's nothing I can do about bad seeing (and we are on a bad site, compared to other big observatories), but if the seeing is too bad, the star gets smeared out so that it is much larger than each fibre, and only a little bit of light will fall down the hole, into our detector. Ironically, if the seeing is too good, the guide star has a good chance of falling through a hole between our 7 guide fibres (which are separated by an arcsec), and will completely miss falling down fibres for the regular sources. It seems the fibre sizes are a poor match for the typical seeing conditions and astrometry errors we expect, but on the other hand, you don't want too much "empty sky" falling down fibres, adding false spectral lines
| | Date : |
Sat Sep 23 21:08:06 2006
| | Event : |
My first week night assisting at the Anglo Australian Telescope
|
|
