Summary
The elevation encoder was replaced last week Tracking through transit revealed elevation errors with the s-curve characteristic, but of varying amplitude - a new, or previously unseen phenomenon. Various hypotheses were tested last week to explain this : ranging from mounting looseness to thermal effects on the anti-rotation rods.

The imbalance of the telescope seems again to be a player in this saga, and so on Monday 30 Apr an additional 30kg was added to the bottom of the telescope. Larger elevation drive currents were then needed to drive the antenna down in elevation. The ratios of currents before and after these weights were added leads us to expect a 1.8x increase in the transit step sizes cf. those seen last week : viz 10" as for a source like Mars transitting at elevation 46 degrees.

Eleven (11) sources were observed either by repeated (SCUBA) POINTINGs or by the map16 method. The nine map16 data are displayed below :

The amplitudes are plotted against cosine(el*evation at transit) below :

The fitted line is forced through the origin, which seems physically acceptable. It has the form

       step size  =    7.5" * cos(el*)  +  0.0"
                    +- 1.9"                1.5"

The amplitude at elevation = 46 (our mars test case) is then typically 5.5".

Conclusion
The relationship is well-defined by these data, even if the amplitude today is not as large as expected.

2001 05 02
A further 30kg was added to the bottom of the telescope during daywork 01 May and a transit track of Uranus performed the next morning :

The amplitude of 4.5" at el*evation=56 is ~0.5" smaller than yesterday, but the difference is probably not significant compared with likely errors. Note also that the Uranus data again shows considerable width, of about 5 degrees, comapred with ~1 degree for the other well-determined values.