Date: Tue, 8 Feb 2000 09:16:04 -0800
From: Russell E Owen
Subject: Secondary position errors (Re: Obs Log 2.5-m (MJD 51582))
The secondary position error will tend to be large after a change in
tilt or after a period of the actuators being idle (as will happen
when the telescope is not tracking).
The issue is that the axial encoders are now Heidenhain incremental
encoders that directly senses mirror position. As a result:
1) The mirrors move around a bit as the grease in the axial actuators
settles down, which will cause the error to increase over time if
motion requests are not received.
2) There are systematic errors during change in tilt due to the fact
that the encoders are not right at the actuators. These systematic
errors can be very large for large tilts (but large tilts are rarely
useful for anything).
The fix is to issue regular position updates to bring down the error.
This happens automatically when the telescope is tracking. The
interval between collimation updates is set in the Tune block. (But
if the interval gets too short there may be too much heat generation
at the secondary.)
In the long run we plan to get the piezos powered up and use those to
reduce the error.
Homing the actuators is not likely to help this particular problem.
-- Russell