Az, Za model offsets and the horn survey.
The receivers are positioned using a theodolite survey.
The telescope pointing model is then made with the sbn receiver. After
the model is made, sources are tracked with sbn (using the new model) and
then retracked with the "other" receiver. The pointing
offsets are then used to compute the az, za offsets for these "other"
receivers. If the receivers had all been positioned correctly, then
the az, za pointing model offsets for all receivers should be the same.
may04 model 15 pointing offsets:
The receivers were surveyed into position and then resurveyed
on 02may04. Pointing model 15 was made after this. The sbn and the lbw
receivers were not moved in the x (za) direction because there was no room
in the hole in the turret floor.
The plot shows
the az, za pointing offsets (.ps) (.pdf)
by receiver for model 15.
To include the survey residual errors for za, I used
13.88 Asecs/cm. This conversion factor comes from 45 asecs
sky/turret degree. The conversion factor I have of 17.04 asecs/cm for the
x direction caused a larger scatter in the green +.
The top plot shows the model 15 azimuth and za offsets for all receivers.
327, 430, and 610 are not surveyed into position. The Black * are the azimuth
offsets (great circle) while the red + are the za offsets.
The bottom plot shows the model 15 az and za offsets for the
receivers that were surveyed into position. black * are azimuth errors,
red + are za errors, and green + are za errors after including the residual
survey errors in za.
After correcting for the residual survey error, the
rms scatter in the za pointing offsets was 1.7 asecs. This corresponds
to about 1 mm rms in linear motion.