Date: Wed, 14 Apr 1999 21:16:29 -0700
From: Russell E Owen
Subject: Another lost night
The servos are still too flaky to use for pointing the telescope.
Problems included:
- Two runaways, one in alt, one in rot
- Immediate offsets often caused axes to stop tracking. This is very
unfortunate, as it greatly hampers the ability to do pointing
modelling. In the long term it is unacceptable. The servos should be
able to handle "+MOVE pos" commands of arbitrary size. It doesn't
have to get there gracefully, but it should get there in one piece.
- Azimuth halted several times for no obvious reason.
- Restarting axes after they halted them was a royal pain, since AXIS
INIT still doesn't fully restore control to the TCC.
- Reported position and TCC requested position showed a steady
disagreement of 0.02 to 0.04 degrees. I'm not sure how this varied
from one slew to the next, but it was pretty steady over time for
tracking. Such an error is grossly too large.
- DRIFT may be misbehaving. Dan saw something funny through the
finder scope while doing small computed offsets (I tried using those
to avoid immediate offsets causing the telescope to stop tracking).
The image would jerk and typically would not settle into the proper
position. My guess is this is an incorrect response to the DRIFT
command, coupled with some kind of hysteresis in the control loop or
mechanics. But it's only a guess. When things are working properly
the sky will start drifting by during the time between the DRIFT and
the new p,v,t triplets, but there will be no abrupt motion.
-- Russell
P.S. I can't say much about the reliability of the rotator. We
disabled it after a slew or two to save time. We figured we could do
without it for the initial crude pointing model.