�� . �� : http://www.naic.edu/~phil/pnt/platMotion/pntErrVsHght.html
�� : Mon Jun 18 23:14:12 2012
�� : Mon Feb 4 10:04:15 2013
��:
�� : phoenix

pntErrVsPlatformMotion

Pointing error when platform moved vertically

jun12

Sections:
The computed error
Comparing the measured and computed error.

Plots:
the computed pointing error for different vertical offsets (.ps) (.pdf):
the measured platform height offset and pointing error (.ps) (.pdf):

Idl routines:

gen/pnt/pnterrplathght.pro
To use:

idl
@phil
@geninit
.compile pnterrplathght
explain,pnterrplathght

The error when moving the platform vertically.

When the platform is moved vertically, the feed moves away from the paraxial surface. This introduces a pointing error. This error occurs during warm weather when the platform sinks below the focus position and there is no tiedown tension left to push it back up.

The Figure below shows the error:

err=dhght*sin(za)/radiusCurvature

When the platform moves up, the za pointing error is positive (as if the dome was at a higher za). To correct for this error, you need to subtract the za Error from the requested za position.
To first order, the azimuth error is not affected by moving the platform vertically (the motion is azimuthally symmetric)

The plots show the computed pointing error for different vertical offsets (.ps) (.pdf):

The different height offsets are coded by color.
The top half of the plots show when the platform is above the focus position.
the bottom half of the plot has platform positions below the focus position.

processing: x101/120615/pntErrVsHght.fig, 120529/pnterrplat.pro

Comparing the computed error with the measured error.

On 15jun12 a calibration run was done using the cband receiver. It consisted of:

It was a few weeks after pointing model18A was installed.. so the pointing should have been pretty good.
Prior to the start of the run, the distomat cpu stopped updating the file containing the platform height.

The online platform height tracking algorithm switched to use the air temperature.
The temperature algorithm is not as accurate as the distomats, so the platform was a little high or low.
The next day we fixed the communications problem, and all of the distomat data from during the calibration run was recovered.
This gave us a dataset of a pointing run with simultaneous info on the vertical error in the platform position

The plots show the measured platform height offset and pointing error (.ps) (.pdf):

Page 1: pointing error vs az,za

Frame 1,2: za error vs za and az
Frame 3,4: az error vs za and az.
black * is B2128+048, Red is source B2337+220
The mean za error is 6.0 asecs
The mean az error is .299 Asecs.

Page 2 height offsets and errors:

Top: The platform height offset vs hour of day.
2nd: The Za position of the two source tracks vs hour of day
3rd: overplot pointing error and computed height error.

* black: measured az pointing error,.
* red: measured za pointing error.
* green: za error computed from the height offset
The height error tracks the za error. It is not correlated with the az error.

bottom: za error with height error removed.

The mean za error is not .032 asecs. This is much closer to what the model predicted.

Summary:

The pointed error computed from the height offset matches the measured za error offsets when the platform is offset in height.
The platform height offset does not affect the azimuth error.
If you have an offset in the tiedown motion:

tiedown motions are 1.7 times larger than the platform motions (cable boom stretch,rotating).

processing: x101/120615/x102cb.pro

<- page up
home_~phil