Much pointing with RxA, in admittedly scrappy weather, has been recorded in the past few days.

    Dataset     N      
  pt020111_17  20
  pt020112_17  17
  pt020114_01  16
  pt020115_16  13

All individual datasets show similar systematic trends of daz-vs-el (upper center) and del-vs-el (center) so they have been combined and are plotted below :

SCUBA data taken during this period showed completely nominal performance :

whereas SCUBA data taken at the end of this period also show the elevation systematic :

Analysing the RxA data with both TPOINT and FIT9 shows the need to change the flexure (TF) of the secondary support from about -13"/90^o to a value of essentially zero, while the later SCUBA data would suggest a new value of about -4". (The SCUBA data from the 14th suggest no change, of course).

FIT9 also allows the determination of possible TMU positional errors for Nasmyth receivers like SCUBA, or cabin flexure for cassegrain receivers like RxA. For the RxA data above a value of u3 (or f1 in the notation of MTIN062, Sect 4.5) of approx 20" is found.

(There are no clock errors in any of these data).

It is unclear at this stage how to proceed. The need to change TF seems fairly convincing from the RxA data, but such a change would be unconnected with any physical event. The ambiguous behaviour of TF in the SCUBA data during this period only serves to confuse.

TSSs and support staff should encourage local pointing while a solution is sought.

18 Jan 2002
Discussion during the JCMT Ops. meeting today suggested that these are two unrelated problems, possibly due to

• a SCUBA fore-optics mirror was 'knocked on 28 Dec. Mirror alignments were checked on the 16th, and will be be repeated 23rd.
• possible loose mirror in RxA optics - HEM to investigate 19th.

Date: Tue, 22 Jan 2002 13:17:00 -1000 (HST)
From: Robin Phillips 

Ken, Peter and I checked the A3 input mirror today 
and found it very definitely too loose. 
The spring that holds the mirror had at some point 
been stretched beyond its elastic limit and was 
therefore no longer holding the mirror as tightly 
as we might have liked. It is possible 
that it was loose enough that the mirror might have 
moved when the telescope was tipped. 
We did a temporary fix to the spring (pending more 
springs being ordered) and then checked the alignment. 
This was slightly out - maybe due to the mirror now 
having a different position due to the adjustments. 
We centred the mm beam on the centre of the tmu and then 
the laser beam on the centre of the tmu. 
The laser beam now carries straight 
on to the centre of the smu.

SCUBA fore optics alignment checked -- OK ? 
(full report by PF soon)

28 Jan 2002
(Blush) The above systematics have been caused by the use, since 18 Dec, of a blatantly incorrect value of TF -- not the usual -13", but more like -26". This value resulted from the analysis of the RxA data taken in poor conditions following the CBLA on 18 Dec 2001 and survived subsequent 'tweaks'. A file-naming error convinced me that I had changed the value at some point in this process, when I hadn't. My assumption that the value of TF extant throughout January was -13 was incorrect, and the data (from both RxA and SCUBA) were indicating precisely that. The SCUBA data of UT20020115 (shown above), are now the only data to *not* show this trend obviously.

I made a bookkeeping error that has generated a lot of work for many concerned with the alignment of instruments and I apologize for the unnecessary work this has caused you.

I have used the SCUBA data from UT20020121 to update the model -- essentially a change only to TF. This should provide pointing free of the gross systematics reported above, but second-order adjustments may still be necessary. Observing staff are still encouraged to do local pointing to mitigate against any (hopefully smaller) residual systematic errors.

05 Feb 2002
Pointing with RxB last night show the new value of TF is working OK.