Each inclinometer was moved to the TMU and oriented with the axis to
be tested perpendicular to the elevation axis. The INCLINE logger task was
started as described above, and the telecope was moved in elevation and allowed
to settle. The stepsize used was initially 1" and the telescope
moved over
a range of a few arcseconds. This yields a trace like that shown in Fig.1.
Steps of 10 arsceconds produced Fig.2
and analysis of the relationship between elevation offset in arcseconds and
inclinometer output in millivolts (Fig.3) yields the scaling factor for
that axis (20.2 +- 0.1 mV/").
The SERVO noise is clearly seen in Fig.1; it has a period of 
3
and
an amplitude of about 1/3", which is the encoder resolution. More accurate
scales may be achieved using faster sampling rates, longer sample times and
larger ranges of elevation, but it is unlikely that the resulting scales will
change much.
However, plans are afoot to rederive the scales to higher accuracy,
and, simultaneously, to quantify the linearity of the inclinometer-encoder
combination.