What's the most accurate polar alignment you should aspire to on an SCT?

[g__day]

I've been driving at polar alignment for quite a while now on a permanently mounted scope on a pier. I have reached the point where I am wondering if its the optical design of a mid size SCT that is restricting me achieving better polar alignment than 1-2 arc minutes.

Basically I am wondering does mirror flop on an SCT restrict one from achieving < 10 arc second polar alignment - becuase what one sees and thinks is DEC drift is actually complex mirror flop.

I'd love to hear folks thoughts on the matter. I have run PEMPro polar align wizard night after night - for 20 - 30 minutes per axis and have drift recorded of under 2 arc secs per 30 minutes - in both RA and DEC. But when I do a 50 star MaxPoint sky model it tells me I am respectively 1 and 2 arc minutes off the SCP.

I ponder is polar alignment error I am seeing modelled actually the result of mirror flop?

Looking at 6 and 8 minute unguided shots at 2.3 metre focal length (a C9.25") I see 1 mm star trails on 1 mm stars - so say a few arc seconds drift - not sure if its RA or DEC.

To prevent the mount tracking fast - with the mount set in polar aligned mode - I have to do a 3 star alignment to get tracking to look right within say 10 arc seconds an hour.

PEMPro also reports tracking PE to be within +/- 2.111 arc seconds raw - so this is kinda superb. So I ponder is it the mount is great but mirror shift / flop on the primary making it impossible for me to get better alignment?

What are your thoughts and how good a polar alignment can you achieve drift aligning with a SCT?

[jase]

When are you going to start imaging Matt?

All this fine tuning means zip if you've got nothing to show.

Do you see field rotation in long guided exposures i.e. field rotates around guide star?

[gary]

Hi,

In my dealings with fellow enthusiasts, one of the most commonly misunderstood
concepts is that of 'polar alignment'.

A common misconception is that by performing a drift test and making the necessary
adjustments to the mount in Azimuth and Altitude, one can align the mount's polar
axis to correspond to the celestial axis giving a result that is 'perfect' for all positions
in the sky.

There are a couple of problems with this. Firstly, the effect of atmospheric refraction
is to make objects appear slightly higher in elevation than than actually are. If this
phenomena simply resulted in a constant amount of shift for all object altitude
positions in the sky, things would be somewhat simpler. However, this is not the
case. Whereas the displacement at the zenith is zero, by the time one reaches
a zenith distance of 45 degrees, objects are displaced by about 1 arc minute
and as one gets closer to the horizon, there is a rapid increase which reaches about
half a degree. The implications of this are that the optimal polar axis varies with
the elevation of the target object. For short exposure times, this is not a problem.
However, unless one has a mount which dynamically shifts the mount in altitude
during exposure, one can end up chasing one's tail looking for that magical
'perfectly polar aligned' position.

Also keep in mind that since 'tracking' is simply the first differential of 'pointing',
in other words, pointing as a function of time, the optimal tracking rate also
continually varies as one goes from horizon to zenith.

The second problem with a drift test is that it is simply providing a polar
alignment estimate based on just the two positions of the sky that
constitute the observational inputs to the test.

Where things become even more complex are due to the effects owing to the various
geometric fabrication errors, bearing errors and gravitational flexures within
the mount/OTA. If the errors are systematic, then one has some hope of
compensating for them by computer modeling. However, if they are unsystematic,
which means they have elements of randomness to them, then they can never
be modeled. A good example of a systematic gravitational flexure error might
be the bending of the forks on a fork mount or the Dec axis on a GEM.
A good example of a unsystematic error might be a sudden mirror flop in an SCT.
As the mirror begins to shift, the movement may at first be relatively systematic
and described by a simple physical model. However, if stiction is involved
which is suddenly overcome and the mirror unpredictably makes a discontinuous
shift, then this is somewhat more problematic and cannot be modeled.

The best available tools for analyzing potential systematic and unsystematic
mount/OTA errors are those that perform an analysis of a large number of
sampled star positions taken across the entire sky. You mentioned you use
MaxPoint. Though this is a fine program that will be suitable for many people,
it is nowhere near as sophisticated as TPoint and for the order of magnitude
of error you are now discussing, you would definitely need to perform a TPoint
analysis. MaxPoint only provides a very limited number of modeling terms
and I know first hand from the type of mount that you are using that there
is a strong likelihood that there will be several additional first order effects that
may improve your whole sky RMS pointing performance even further. I know
this from our own work with the Argo Navis Telescope Pointing Analysis System
(TPAS) which took us many years to devise, implement and test. Though
TPAS does not provide as large a suite of terms as TPoint, it does
include some of these often very important additional first order effects.

The advantage of analysis tools such as TPOINT, TPAS and MaxPoint is that
they provide a polar alignment estimate based on a large number of sampled
positions rather than just two positions, as with a drift test and that they simultaneously
take into account geometric and mount/OTA flexure errors.

When you look at the result from PEMPro and the result from MaxPoint and with the
above discussion in mind, you might now see these results in a new light.
Whereas your MaxPoint result is providing a polar alignment correction based
on the 50 star sampling run across the sky, your PEMPro result will be based on
30 minute run of whatever part of the sky you happen to be pointing to at that
time. However, you should continue to use both types of tools.

When you mention the order of magnitude of error as in the 1 to 2 arc minute
range and given you are only sampling 50 stars and then using MaxPoint rather
than TPoint for analysis, this also tends to be a flag for me that you are possibly
also starting to run up against the limitations of such a modest sampling run
and the inherent limitations of the MaxPoint tool. For the obsessive or curious,
I would recommend considering a run of 100 to 150 stars and using TPoint to help
glean out any additional first order and possibly second order and higher harmonic
effects. Anecdotally, based on our own field work combined with our professional
knowledge of mount error analysis, one typically needs this order of sampling
in order to reliably distinguish often subtle flexures from the rest of the noise.
Thankfully the polar mis-alignment terms have relatively distinctive signatures
and tend to relatively well characterized. It is recommend you align the
mount's polar axis with the refracted pole rather than the true pole.

As for a mirror shift, this is absolutely possible and is not uncommon on large
SCT's, such as 11" to 16" apertures. In the end, what you are trying to achieve is the
elimination of the star trails. One possible approach would be to perform a long
sampling run and consider familiarizing yourself with the graphing functions
of TPoint. Keep in mind that TPoint was originally designed for professional
practitioners and in order to use it to best effect, one needs to carefully consider
how to hypothesize the right question in order to get the information one is seeking.
Specifically, one needs to think about what the two axes of the graph should be
in order to see where there is a sign of some unmodeled effect. For example,
though it is a term normally used only on Alt/Az mounts, one might consider
some plots with the IE term (Index Error in Elevation) on your polar mount
as one candidate that might help you catch any mirror flop with its pants down.

Best Regards

Gary Kopff
Managing Director
Wildcard Innovations Pty. Ltd.
20 Kilmory Place, Mount Kuring-Gai
NSW. 2080. Australia
Phone +61-2-9457-9049
Fax +61-2-9457-9593
sales@wildcard-innovations.com.au
http://www.wildcard-innovations.com.au

[g__day]

Jase, Gary - thanks for your thoughts!

Yes - I am going insane with this - plus cloud cover has been about 76 of the last 80 days!

I did Ghost of Jupiter last night - short shots and was very pleased with the result. Move onto M83 at 6-8 minutes exposures unguided and I saw slight star trails!

The mount (a Vixen Atlux on a large pier) is good enough - its hand controller at 130 pages is just not well documented enough!

Gary - MaxPoint can hold many more than 50 stars - you can model on hundreds of stars using your own sky control program and just manually adding adjustment points; but yes I am beginning to realise it doesn't have the smarts of Tpoint - which I will likely buy in the comming months.

Also the SkySensor2000-PC models and corrects for the airs refraction and /or the different motion rates of stars, planets, the moon and satelites if you program the last. But it's barely even documented! I don't know how well it does it!

I realise too that I started at the real deep end - imaging at long focal length rather than on an 80mm apo at 500mm - 800mm. Call me a machoist!

To I haven't been able to get all teh weak points out of my gear yet, some examples:

1. I walk around my pier - if I stand on it at the South West side - stars move 20 arc seconds or more - bolting it through pavers onto a large concret block was a truly dump idea! I should put bolts on hard against the pavers themselves, rest the base of the mount on these four bolts & washers, then lock bolt it on top. That would be much more secure!

2. There is small slop in DEC - I can re-position the worm a bit closer and get better pointing and alignment.

3. I will add a feathertouch fine focuser and eventually buy a good apo of reasonable size - I think I just missed out what I wanted by a day in fact!

4. I will add Tpoint and have asked Ray Gralak to change PEMPro's polar alignment wizard to run via MaximDL with a minimum shot duration of 1/1,000 of a second (present minimum is 1 second) - so I could polar align during the day (Ray says this is an easy fix).

5. I will keep on trying to find out what the two numeric compensation parameters of the Atlux in Polar aligned mode are - and how they affect tracking and pointing.

I will practise, practise, practise and eventually get a better (likely self guiding S-BIG camera) and/or do on axis guiding with a giant Lumicon on axis guider.

But bottom line I want some fun now - there has been alot of set-up, now I want it to work. I want to be doing tuned imaging soon!

Matthew