August 21, 1996
News from UKIRT

Dear colleagues,

After the immediate success when switching on the tip-tilt system on Wednesday night (Aug 14, the last of the UKIRT engineering nights), we started to fine-tune the system during our first three nights (Aug 15-17).

The general message is, that the tip-tilt system works excellently. In the best images the FWHM was taken down from an uncorrected 0.5'' to a tip-tilt corrected 0.25'' and the peak intensity tripled (see Fig.1). Over more than an hour the corrected FWHM was below 0.3''

We also tested the limiting magnitude by putting neutral density filters into the optical path of the wavefront sensor. It is a bit difficult to give a hard number here, as the old dichroic is still in place. According to Nick Rees, the stars will appear two magnitudes brighter with the new dichroic that was produced by Zeiss using the same design as for CHARM and ALFA, than they appear with the old dichroic. Without going into too much detail, one can say that starting with a flux equivalent to an 11.5 magnitude star that gave the best image quality, a good improvement was still achieved with a 14.5 magnitude star, and we think that only at 16th magnitude there will be no more improvement in the FWHM. This is the limit that we intended to achieve. This is especially good news for us as the same hardware and software is used in CHARM and ALFA.

Tip-tilt correction on both ends of the chop works too (see Fig.2). Here, the FWHM was improved from 0.69'' to 0.3''. I think, this is the only adaptive optics system that can do tip-tilt correction on the chop ends. Hopefully, Tom Herbst and Massimo Robberto will make good use of this feature in the next week.

Another important test was to point the telescope into a 40mph wind. Without correction, the star image was egg shaped, with correction it was a nice round image with a peak intensity double that of the uncorrected star. The FWHM went down from 0.59'' to 0.28'' (see Fig.3).

The original worries about the mechanical oscillations of the telescope around 20Hz have largely disappeared due to the excellent image quality that we achieved. The analysis of the image motion shows that the rms value of the image centroid motion is reduced typically from 0.23'' to about 0.035'' (see Fig.4). This has to be compared to a value of typically 0.025'' rms image motion that we achieve with CHARM using the 3.5m telescope at Calar Alto that does not have the high freqency oscillations. In the power spectra, one can see an resonance frequency at 0.8 Hz (see Fig.5) that is caused by the telescope encoder.

This is the status after half the run is over. We will do more tests on the adaptive focus correction and check the performance when correcting with a reduced aperture to take out the coma. Last but not least, we will also do some astronomical observations.

It has to be mentioned here, that it was only possible to achieve 0.25'' images because Tim Hawarden, Nick Rees, Tim Chuter and Chas Cavedoni did an excellent job in minimising the optical aberrations of the telescope by using a static wavefront sensor. Thus, astigmatic and trefoil aberrations could be corrected by adjusting the primary mirror support system. It might be a good idea to do the same kind of thorough analysis at our 3.5m telescope that has suffered from aberrations for at least two years now.

Viele Grüsse

Andreas Glindemann