Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.naic.edu/~phil/hardware/vertex/azencoder1fail.html
Äàòà èçìåíåíèÿ: Thu Aug 6 23:07:10 2009
Äàòà èíäåêñèðîâàíèÿ: Tue Aug 18 09:04:51 2009
Êîäèðîâêà:
Ïîèñêîâûå ñëîâà: http www.badastronomy.com bad tv foxapollo.html

az encoder1 fail

Azimuth encoder1 failures

aug09

    There are two azimuth encoders. Encoder 1 is on the gregorian side and is used for pointing. Encoder 2 is on the carriage house side and is used to measure azimuth pending (as well as a check on the gregorian side encoder).

    When encoder 1 fails, the vertex system automatically switches to encoder to for the pointing. This causes a degradation in the pointing since the encoder 1 and encoder 2 values vary as a function of azimuth. This is caused by the "out of roundness" of the azimuth encoder rack gear as well as the bending of the azimuth arm.
    After vertex system will continue to use encoder 2 for pointing until the encoder 1 error is cleared. The error can be cleared by:

On the vme system:

pnt reset 1.. this will clear the error without stopping the tracking (unless a motion failure occurs).
pntpwrdip : this will stop the tracking, clear the fault and then start tracking again.

From the ocu:

Going to the faults window, then azimuth, the pushing the clear key.

The telescope may get a motion failure and stop tracking during the reset if the difference between encoder 2 and encoder 1 are large enough (the software sees a large jump in the position and then stops). In this case pntpwrdip will try to restart the tracking.

The plots show a jump in the azimuth position while clearing an encoder1 fault (.ps) (.pdf):

The telescope had been sitting at az=285, za=12 when the failure occurred. It was still sitting there when the failure was cleared.
Page 1:

Top: shows the azimuth status vs time. The encoder 1 failure is the lower red trace. It was cleared at 11:35 and again at 11:57

Bottom: shows the azimuth position returned from the vertex system during this time. The azimuth is jumping by about .01 degrees. After the failure occurs, the new position stays at encoder 2 until the failure is reset.

Page 2: how large a pointing error can a failure introduce?

6 azimuth swings (-90 to 270) were done on 02aug09 (looking for rfi).
The top plots shows the encoder difference (encoder 1 - encoder) in little circle azimuth degrees. The jump that occurred at 11am was at an azimuth of 285 (-80). The difference plot shows this to be about .01 azimuth degrees which matches the jump we saw.
bottom plot: This is the pointing error on the sky for an encoder jump.

the sky error is sin(za)*azencoder jump\
black is at za=12
red is at za=19
the error gets up to about -65 asecs (great circle). The encoder difference is not currently centered at 0.

Summary:

Encoder 1 failures cause the system to use encoder 2 for pointing.
The system remains with encoder 2 until the fault is reset.
The pointing error can get up to about 65 Asecs (great circle) depending on the za and the azimuth position.
See also: enc1failures aug09.

processing: x101/090802/enc1fail.pro

home_~phil