Paramount ME User’s Guide
87
ME - Error in polar axis elevation
MA - Error in polar axis azimuth
Do not just go to the Suggest Terms command that allows TPoint add terms;
these are merely suggested terms, and must be analyzed and evaluated before
you can trust them. That's very challenging to do. There's a better way. Add,
then evaluate for effect on the model, the following terms:
TX – Tangent tube flexure. This is generally better to use than TF; it
combines for greater accuracy with the other terms.
HCES and HCEC – Centering errors in the RA gear.
DCES and DCEC – Centering errors in the Dec gear.
DAF – Sag of a cantilevered declination axis.
The above terms are almost always beneficial. Next step is to evaluate the other
named terms (except those with AAT in the name; those are for the Anglo-
Australian telescope, not the Paramount ME). Look for terms that, when used,
will significantly reduce the Sky RMS and PSD values. (RMS = root mean square
error; PSD = population standard deviation.) Never use a term that increases the
PSD, even if it reduces the RMS pointing value. Terms that provide only minimal
improvement (less than 10% better RMS) should not be used. Terms that do not
provide at least 2-3 units improvement in RMS should not be used. Other useful
terms include: HDSD4, HHCH4, and HHSH4.
At this point, you can now try Suggest terms. Use the same rules for suggested
terms as for named terms: make sure they show enough of an improvement in
your point to be included. However, when suggesting terms, you should always
take an additional step: test the term in practice. Suggested terms can be very
useful; they can be completely bogus. The name says it all: these are
_suggested_ terms. They have been found to be of _possible_ statistical
significance; you need to verify that by seeing how they affect real pointing, not
just the RMS and PSD. Having a low RMS value does not necessarily equate to
accurate pointing.
Note: Adding and evaluating terms is a non-trivial exercise. The more you know about
mathematics, and statistics in particular, the more effective you can be at selecting,
analyzing, and evaluating terms. Read up on TPoint, the mathematics and the concepts
behind telescope modeling – it will make you more effective at selecting terms. Also, be
patient. It takes time and effort to learn TPoint and the vast logic behind it. The good
news is that the more time you invest in TPoint, the more you will get out of it.
If you experience problems with ProTrack, the first place to look is at the terms you added
to your model. Terms must be evaluated by how they improve or hurt your pointing
and/or tracking accuracy. The next step is to verify that you cleaned out bad data from
the list of points. Look for outliers and mistakes, and mark them so they are ignored.
Next, be aware that when you are closer to the edges of the model, modeling is less
accurate. If you want to image at 45 degrees of elevation with ProTrack, for example, go
well below that when mapping points. How low you can go depends on the point where
atmospheric refraction begins to elongate stationary stars for your focal length/pixel size
combination.
To Next Page
Loading...
