Документ взят из кэша поисковой машины. Адрес
оригинального документа
: http://www.naic.edu/~phil/pnt/pntchk_nov06.html
Дата изменения: Wed Nov 29 02:06:43 2006
Дата индексирования: Tue Oct 2 05:26:20 2012
Кодировка:
Поисковые слова: mdi
The azimuth 1 encoder failed on 14nov06
and had to be reset using the index on the azimuth arm. X102 runs were
taken on 16nov06 (lbw) , 23/25nov06 (sbn) , and on 26nov06 (lbw). The results
of these pointing measurements are shown below.
Page 1 the top 2 plots show the za Error and az Error versus azimuth.
The sources are color coded. The bottom 3 plots show the tensions in tiedowns
12,4,8 while these measurements were being made. Tiedown 8 lost tension
around az=100 with B1140+223 and at az=50 with source B1317+179. These
points are not included in the following plots.
Page 2 Pointing errors Vs az,za The top two plots show za error
vs za and az. The bottom two plots show the az error vs za and az. The
black + B0340+048 have a large az error offset. You can see a 1 azimuth
term in the za error vs az and a za slope in the za error vs za. The pointing
errors are:
ZaError: offset=1.41 Asecs, rms: 7.65 Asecs
AzError: offset=7.08 Asecs, rms: 7.90 Asecs
Page 3 Pointing errors color coded by receiver These are the same
plots as page 2 except they are now color coded by receiver (lbw black,
sbn red). The za dependence of the azimuth errors is more pronounced for
the lbw (black) points than the sbn (red).
Page 4 1az and 1az, sin(za) fits to the errors. A 1 azimuth term
in the pointing errors can occur if the platform tilt changes from the
value it had when the pointing model was made. This could occur after one
of the tiedown encoder failures (if the new encoder was not set to the
correct position). If the azimuth encoder was not reset correctly then
there would be an azimuth error times the sine of the zenith angle (since
the reported errors are on the sky). To test for this i fit the following
models to the za and az errors:
zaError= C0 + C1*cos(az) + C2*sin(az)
The top plot shows the 1 az fit to the za errors. The amplitude is
9.5 Asecs and the phase angle (of peak) is 45.4 degrees. The fit rms is
2.9 Asecs.
The second plot shows the 1 az fit to the azimuth errors. Both the
data and the fit have had the sin(za) portion of the fit removed. The amplitude
is 7.1 Asecs and the phase angle (of max) is -59.8. This azimuth fit has
an rms of 4.9 Asecs. A platform tilt should have the maximum error for
azimuth and za out of phase by 90 Degrees. The measured difference is 105
degrees.
Sin(za) portion of az error fit.
The bottom plot has the sin(za) fit to the azimuth errors (after removing
the 1 az term from the fit and the data). The amplitude is 45.9 asecs (at
the azimuth encoder). The radius of the encoder rack gear is about 60 feet.
This translates to an azimuth error in the positioning of about .19 inches.
At 10 deg za this would be a 7.8 Asec error on the sky (about equal to
the azimuth offset error of 7.08 Asecs).
There is a 1 azimuth term in the az/za pointing error. It's amplitude is
9.5 Asecs with a peak at az=45.4 degrees (I'm using the za error values
since the fit was much better). It is probably coming from the tiedown
cable lengths changing from the values when the model was made.
A sin(za) term in the az errors of 45.9 Asecs exists. This would be an
error at the encoder of about .19 inches. This error is more prominent
in the lbw data than the sbn data. At 10 degrees za this is an error of
7.8 Asecs on the sky. The measured average az offset was 7.08 Asecs.
The measured pointing error needs to be subtracted from the computed az,
za value to point at the source. A positive za Error means the dome needs
to be brought down. This can be done by pulling on a tiedown. Using the
za Error of 9.5 Asecs and azimuth of 45.4 for the peak Za error, the tiedowns
need to be moved by [.094, .027,-.122] inches (td12,4,8). Positive numbers
mean the tiedowns need to pull the platform down. The reference position
for the tiedowns at 70 deg F:
14.969, 14.643, 13.122 inches (new reference position)
I need to do azimuth spins with the tilt sensors to measure the tilt of
the platform. From those values we can compute the new values for the tiedown
cables to level the platform and compare them with the pointing errors.