Because of the proximity of SPIFI to the elevation encoder a reduced plan
of maintenance is to be performed today (19 April):
Date: Thu, 19 Apr 2001 09:20:40 -1000
From: Dean Shutt
Subject: Details of today's elevation encoder work
Confirming our discussion of this morning:
We will not remove the elevation encoder today. There simply is not
enough room to do this properly with SPIFI in place. However, Vernon
and Tomas are enroute to the summit to perform the following work:
1. Tighten the eight bolts holding one side of the encoder coupling.
These can be accessed through the hole in the elevation bearing.
2. Tighten the encoder anti-rotatation pin. The encoder is held in
place circumferentially by a spring loaded bolt which bears on a
plate. Should this device be slightly loose, or the spring broken
or crushed, the encoder would have a slight circumferential play.
Note that at the diameter of the encoder only approx. 1 micron of
circumferential play is required to cause a 1 arc second change in
elevation on the sky.
3. Inspect the coarse encoder gear and elevation overspeed sensor for
freedom. Abnormal friction in this area would also contribute to
the observed effect.
Having accomplished the above, we will have addressed all the likely
problems in this area excepting the bolts on the blind side of the
encoder flange. Tightening of these requires removal of the encoder as
originally proposed.
I would ask that the transit tracking test be performed at the most
convenient time between now and Monday morning. If the S effect has
disappeared or been significantly reduced I suggest we do nothing
further immediately. We can remove the residual effect in software and
see if the required correction changes over time. If the testing
reveals the S effect is still large we will remove the encoder assembly
for inspection and tightening of the blind coupling bolts on Monday. If
THAT does not resolve the problem, we will think harder ...
As discussed, the change in shape, amplitude, and zero crossing of the S
correction over time is characteristic of a frictional effect. For this
reason, and because the measured changes in response bear no
relationship to known elevation balance changes, I do not believe we
should pursue the "unbalance windup" theory until after completion of
the above work.
Dean
14:00 HST : Vernon reports visual inspection completed, no obvious
motion of bolts during attempts to tighten them, but that the encoder
anti-rotation rods were tightened (item 2 above).
19 April 2001
An elevation encoder zero-point change of 133" resulted from the
engineering today. After the pointing model was updated accordingly,
a further transit tracking test was performed on 3c279, with the
empirical correction disabled, as follows :
- from (az,el) = (158,63) at HST 22:43 to (222,57) at 00:51.
- transit occurred at HST 23:24 at an elevation of 64 degrees.
The seeing was about 0.5", tau_cso ~ 0.05.
The pointing residuals (daz in white, del in red) are displayed below :
The green line is the fit determined form the Mars data earlier in the
day, and the green dots are the residuals of the new data about the green
line. The performance is essentially indistinguishable from that before
the engineering. Even the excursions in both daz & del at azimuth 215
are similar to those seen in the earlier data : although of secondary
concern they suggest that more extensive runs will be necessary to
fully characterise the tracking behaviour.
20 April 2001
The elevation encoder is scheduled to be removed and replaced with a spare
on Monday 23 April, with afternoon/evening checks on the zero-point and
transit tracking characteristics.
Iain Coulson
Latest Update : 20 Apr 2001
