WEL & IMC arrived 21:30 : JGW had just recently opened up and had done some noise measures and skydips. Initial pointing & focussing done on Uranus. For one hour from ~HST 22:00 Uranus was 'tracked' using the the map16 method from (az,el) = (184,56) to (209,52). Tau_cso was about 0.14 throughout. The pointing residuals are shown plotted against azimuth below :

Systematics are small (<2") and are as likely to be due to sky variations as to track or pointing errors. Formal RMS residuals in (az,el) are (1.0",0.8"). We have no seeing monitor data to compare with, but I wouldn't be surprised if a large part of this scatter is due to seeing.

So : tracking in the south seems OK - shame we missed southern transit.

Between 23:00 and 00:00 we did some pointing, mainly in the north. A large -ve change to del in the north cf on Uranus in the SW prompted a tracking experiment through northern transit. After a check on the focus, we tracked 0133+476 from HST 00:12 to 03:00, from (az,el) = (387,57) to (333,57). The residuals are shown below :

Both azimuth and elevation pointing residuals show a strong tendency to be -ve in q5 (az > 360) cf in q4. The behaviour across this azimuth range is mainly monotonic, but with occasional sharp excursions of amplitude -3" or -4", particularly noticeable in azimuth at azimuiths 346 and 373. These azimuths are where wheel 2 (right rear) rolls over joints 8/9 and 9/10. There has been no indication from recent inclinometry of anything that would explain these excursions in the azimuth residuals. By contrast, neither the step in azimuth residuals at azimuth 367 nor the 2" excursions in elevation residuals at azimuths 350 and 360 have corresponding track features.

The obvious systematics here force a rethink on the structures seen in the earlier tracking experiment - even if they are of a lower amplitude and are free of sharp, joint-related features.

Allsky pointing was also done in between all this. Residuals plotted against azimuth, elevation and time are shown below :

Residuals in q5 again seem a couple of arcseconds more -ve than those elsewhere, although there are other non-random features that appear equally suspicious.

Tau_cso remained > 0.12 throughout almost all of this night.

Conclusions
The overall performance seems unaffected by the AEC PLC upgrade and is typical of that seen in recent times of poor-ish weather. (Moderate and encouraging improvement in pointing was recorded during the recent spell of grade 1-2 weather). The q5 systematics :

• -ve residuals in both daz and del
• track-joint related excursions of amplitude 3-4"