Drift alignment corrections - use angular size of Jupiter as a judge?

[g__day (Matthew)]

Folk I was wondering if this makes sense. If I drift align and calaculate how far an axis is from the refracted pole, can one simply use Jupiter's angular (equatorial) diameter of approx 50 arc minutes 2011/2012 as a good visual indicator of the correction amount?

Borrowing a formulea

Alignment Error (arc seconds) = arc tan(drift in arc seconds/(15*drift in time seconds)*360)/(2*PI) * 3600, e.g.

Drift arc seconds = 25 arc seconds (say 1/2 Jupiter's apparent diameter)
Drift in time = 3600 seconds
Polar misalignment error (arc seconds) =95.5 (a bit under two Jupiter diameters)
or Polar misalignment error (arc minutes) = 1.6



So if I centred the scope on Jupiter and it drifted to the edge - the scope was wandered about 1/2 of 50 arc seconds ~ 25 arc seconds. Let's say it took an hour to drift in DEC from Jupiter's centre to its edge - then from the above formulea I am 95.5 arc seconds or 1.6 arc minutes off the pole.

Expressed another way in that example I am just off twice Jupiter's diameter off the pole. So if I change the Alt or Az (as required) - using's Juptiers diameter as a guide stick - then I'd have a pretty neat frame of reference to polar align with.

Does this make sense to others - I have been searching for a visual means to guide me with what an arc minute or two looks like in my main OTA. Does approximating one Juptier diameter is 50 arc seconds at this time of year provide a simple measuring stick for fine tuning polar alignment?

Matt

[PRejto (Peter)]

Matt, I'm sorry nobody has taken your challenge to respond to your interesting thought problem! I don't know enough about this to feel any confidence in commenting, but I do have a question. Does Jupiter really move at sidereal rate? I assume not really, but does it move differently enough to throw off your calculations, especially if you are tracking for hours to measure a small change. Perhaps there is a way to compensate for the non-sidereal rate if that is a factor.

[g__day (Matthew)]

Hey Peter,

I should have been clearer! I don't want to drift align using Jupiter, I want to drift align normally but manually calculate what the compensation adjustment should be - then use Jupiter's diameter as a visual reference to get the magnitude of adjustment correct.

So the drift alignment session might go like this.

1. Note how big Jupiter appears in my imaging camera's view finder - it fills up half the central square focusing box - so the box itself must be around 100 arc seconds corner to corner - that's my imaging guide.
2. Guide on say Rigel when its between 25 - 30 degrees elevation for 30 minutes.
3. Either using the eye piece or a before after shot not only notice the direction of the DEC drift - but measure it manually to get a magnitude number in arc seconds.
4. Plug that drift magnitude and time into the above formulea and note its magnitude of correction. Lets say it was lower mount 100 arc seconds.
5. Goto Jupiter and centre it on a corner of my imaging camera's central box.
6. Manually lower my mount 100 arc seconds by thinking of it as two Jupiter diameters.

So really the change I am suggesting to a normal drift alignment is to:
1. observe the DEC drift angle for a given time
2. manually calculate the polar correction required
3. execute the polar axes correction using Jupiter's diameter as your angular yardstick

The alternate to this is to either simply guess how much correction is required or use software like PEMPro's Polar alignment wizard (which is throwing very wierd correction numbers at me - so I want to manually confirm how much polar misalignment I may have).

So lets say I track Rigel for an hour and I measure its drifted in DEC say Half of Jupiter's diameter in my viewfinder. If I know drift is close to 25 arc seconds in 3600 seconds, the the formulea tells me the correction magnitude is ~ 96 arc seconds - or about two Jupiter diameters. I want to use Juptier as my angular measure to make the physical adjustment of my gear.

Does this make more sense?

BTW - I think my mount actually tracks planets at non sidereal rates, but that's another variable I don't want to bring into the mix.

Cheers,

Matt

[bmitchell82 (Brendan)]

While its a good little experiment and fun, why re invent the wheel? drift alignment does it accurately and software will have you polar aligned in minutes to the arc second accuracy? just my thoughts.

Brendan

[g__day (Matthew)]

Brendan,

Something wierd is happening for me - Tpoint, MaxPoint and PEMPro all disgree. When I do a 3 star alignment and go to the third star pointing is off by about 2-3 arc minutes - maybe its cone error, maybe its alignment error. I tend to think if the error in my 3rd alignment star is about 2-3 arc minutes that should also be my polar misalignmnet error.

MaxPoint concurs with this, but Tpoint disagrees and doubles the error margin and says its in opposite directions (both run on 120 evenly distributed stars)! One run of PEMPro says spot on another a day later says adjust by 15+ arc minutes - so I don't understand why I am seeing so much variability on a permanently, pier mounted setup!

I tracked Jupiter a few days ago and saw about 30 arc seconds drift in over an hour - that's when I remembered the drift formulea to double check my results. Regardless when professional level software disagrees - I put it down to user error and try and gather more data before I adjust!

BTW - is my thinking that once I align on two stars the third alignment star's misalignment should be a solid indication of the sum of mis-alignment and cone errors?

Matt

[originaltrilogy (Petr)]

I use alignmaster and let computer do all hard thinking for me!

[g__day (Matthew)]

Petr,

A few questions:

1. How accurately (arc seconds) would you estimate alignmaster can position your mount?
2. Does it work on the SCP or the refracted SCP?
3. If it only does a two star alignment how does it account for cone error?

Matt

[bmitchell82 (Brendan)]

well i get reports back saying that im within a minute of the pole and my slews confirm that as with just one sync, it will do meridian flips and land the target within 5 min of my center of view. not bad with no pointing model, flex adjustment, refracted positions etc etc.

[g__day (Matthew)]

That's pretty good! Meanwhile I feel like I'm on a quest for the holy grail - trying to eliminate that last bit of polar misalignment! For once I'm wish my handcontroller didn't remove polar misalignment when it tracked.

A question I've asked the Skysensor2000 Yahoo group - that has gone unasnwered so far - I'd be keen if anyone here has the answer or strong views:

If you wanted to drift align – and ensure the SS2K isn’t doing anything fancy would you set it into:

1. Polar aligned mode or polar unaligned mode – and why?
2. Align it on 0 (unaligned), 1, 2 or 3 stars – and why?

I am thinking of going with Polar aligned with 2 stars – I’d be interested in others peoples thoughts and experience!


Thanks,

Matt

[gary]

Hi Matt,

Before I respond to this thread I must admit I had been secretly hoping that someone
who had attended my presentation at the recent IceInSpace Astro Camp on
"Busting the Myth of the Perfect Polar Alignment" may have chimed in.

Alas, as I feared, they must have all fallen asleep.

Originally Posted by bmitchell

drift alignment does it accurately and software will have you polar aligned in minutes to the arc second accuracy?

Brendan, I sense from the question mark at the end of the sentence that you
may be uncertain as to the statement's validity. No problem in making an educated
guess.

Unfortunately it is not entirely correct.

However, if anyone who happened to attend my talk would care to politely explain
why it is not correct, I will then know I got the message out to at least one person!
We raised this very point in the presentation, so any takers?

Best Regards

Gary Kopff
20 Kilmory Place, Mount Kuring-Gai
NSW. 2080. Australia

[gary]

Hi Matt,

Despite the fact we want to encourage people to come to IISAC for the dark skies
and hopefully for the presentations, I don't want to leave you completely dangling.

Some time back when you posted your pointing data I had a quick look at the time
and with a busy schedule, didn't get a chance to follow-up. Though I have forgotten
the specific details, my recollection was there were indications in the data of
hysteresis of a similar order of magnitude of the discrepancies in your results.
In other words, it looked as if you were coming up against the finite amounts of
backlash and other random errors within the mount beyond which you would be
unable to improve the pointing further without doing something physically to the
mount's mechanics.

The bottom line is that the data was indicating that due to these random errors, you may
not be able to appreciably reduce the zone of uncertainty in your alignment with the true
pole beyond what you currently have determined it, unless you further minimize the random error sources.

Best Regards

Gary

[bmitchell82 (Brendan)]

the ? was a accident Gary, at the end of the day you are dead on with what you have said I cannot dispute that. but I know what i see when its landing my targets virtually on the middle of the CCD sensor with a 50x40minute FOV at 1200mm with 1 sync in EQMod do you not agree that this is good enough even in astrophotography terms?

Im happy to reccomend to anybody using the software. In regards to mechanical alignment, cone errors and the like we could sit here for years getting it perfect but thats a futile and pointless escupade. So as a yard stick when alignmaster tells me that im within 1 minute of the pole i don't sit there for another 3 hrs trying to get zero because i have back lash in the mounts drives and at one minute of arc bugger thats close enough for the guider to suck out mind you i get to that accuracy after the second iteration which will generally be after 5 minutes. bang for buck my 16 dollars was well spent.! Don't take what i say the wrong way, typing doesn't show my inquisitive nature and if im wrong please im not precious say so as i might be reading and mis-understanding what your saying.

Brendan

I would have loved to have been at IISAC but unfortunately West to East coast in the middle of exams isn't possible. Ill make it one day though!

[gary]

Hi Brendan,

Thanks for the post.

Originally Posted by bmitchell82

the ? was a accident Gary, at the end of the day you are dead on with what you have said I cannot dispute that. but I know what i see when its landing my targets virtually on the middle of the CCD sensor with a 50x40minute FOV at 1200mm with 1 sync in EQMod do you not agree that this is good enough even in astrophotography terms?

Absolutely and this was also a key point of the presentation of Lostock.
In other words, the end very much justifies the means which is really the
key point of your original response.

Im happy to reccomend to anybody using the software. In regards to mechanical alignment, cone errors and the like we could sit here for years getting it perfect but thats a futile and pointless escupade. So as a yard stick when alignmaster tells me that im within 1 minute of the pole i don't sit there for another 3 hrs trying to get zero because i have back lash in the mounts drives and at one minute of arc bugger thats close enough for the guider to suck out mind you i get to that accuracy after the second iteration which will generally be after 5 minutes. bang for buck my 16 dollars was well spent.! Don't take what i say the wrong way, typing doesn't show my inquisitive nature and if im wrong please im not precious say so as i might be reading and mis-understanding what your saying.

Great stuff!

Perhaps another way to express essentially the same thing might be to say
something like "the method and tools by which I do my alignment cost little in
terms of time and money and when combined with keeping imaging times short,
the FOV not too wide, having autoguding and variable tracking rates
and by choosing optimal regions of the sky, the end result has no discernible field rotation".

Appreciate not everyone can make it to Lostock.

I was having a little fun when I was hopeful there might have been one respondent
to the thread who might have said, "I went to this talk by Gary at Lostock
and he talked about some of this polar alignment and refraction stuff even including
its history with regards astrophotography dating back to 1890".

I'll wait a little bit to see if anyone chimes in (I hope so!) and in any case
I promise a follow-up with regards the sentence I said was unfortunately
not entirely correct. Please understand it is not my intention to offend
you and again in the talk I mentioned that many of the subtleties of these
concepts are very commonly misunderstood as evidenced on countless
threads on many forums including right here on IIS. If there was a mission
statement to my talk it was to try and help bust the commonly held notion that
there is some magic point in the sky to which one can align the polar axis
of the mount which is then universally ideal for all imaging.

I would have loved to have been at IISAC but unfortunately West to East coast in the middle of exams isn't possible. Ill make it one day though!

Hope they went well and it would be great to see you over here one year!

Best Regards

Gary

[g__day (Matthew)]

Gary

Out of interest did your talk cover what path a star will travel through the sky? I expect most folk would say a circle, then change that on reflection to be an elipse (unless you're at the equator). But on further reflection - once you count in altitude and elevation based refraction - the path of a star across our night sky is likely to be a hyperbola.

So if a mount's controller realises that and models it correctly (mine uses King rates for refraction to compensate for this - as a function of elevation) how close to perfectly aligned can one come?

What I am saying is give a mathematican any spot on the Earth, including elevation - surely they could model the position of any star across the night sky down to the sub arc second if required (so long as they know the density of air above them and can ignore turbulence or large changes in air pressure systems). Once one factors in turbulence I imagine seeing can move a star +/- 1-2 arc seconds at sea level - does that sound right?


My mount did seems to show some hystersis - which I put down to slop in the DEC gears - probably about 15 arc seconds as a guess. But I don't see that this is the x-factor stymming my polar alignment when I am only watching the long term movement of a single star. I agree other factors such as the non orthorganality of any component in the mount / OTA system can cause pointing errors.

Cheers,

Matthew

[originaltrilogy (Petr)]

Originally Posted by g__day

Petr,

A few questions:

1. How accurately (arc seconds) would you estimate alignmaster can position your mount?
2. Does it work on the SCP or the refracted SCP?
3. If it only does a two star alignment how does it account for cone error?

Matt

Within 1 arcsecond

[bmitchell82 (Brendan)]

ide have to say petr unless you have a really really high quality mount and there are no errors in your system, align master will not give you 1 arc second accuracy it isn't really possible and hence why Gary posted. As a user of alignmaster ide estimate it to be +- 1 arc minute at best which is still more than sufficient!

[Terry B]

I just use guidemaster. It has a drift align mode. As I have a permanent setup I only did it once and check it a few times per year. I don't care what T point states the error is as long as I have no drift in dec. If there is no descernable drift in the time of your longest exposure then that is good enough.

For the SS2000 the recommended mode for drift align is polar aligned and a 1 star align. If you use more than that it is too clever and will try to correct the alignment error it measured and will possibly bump the dec axis. This defeats the purpose of the drift align routine.

[g__day (Matthew)]

Originally Posted by Terry B

For the SS2000 the recommended mode for drift align is polar aligned and a 1 star align. If you use more than that it is too clever and will try to correct the alignment error it measured and will possibly bump the dec axis. This defeats the purpose of the drift align routine.

Terry - huge thanks for that - it's exactly what I needed to hear. As soon as the skies clear I'll try again.

BTW - I do have PoleAlignMax and the full DC3 pinpoint version - haven't experiment with them yet. I think I'll try a manual drift align - plugging my data into that little arc tangent function to see what drift corrections it advises then using Jupiter's diameter as my measuring post. I'll also give PEMPro another whirl now I know to use the SS2K in one star - polar aligned mode!

Terry out of interest - did you read that somewhere or were you able to deduce it from experience?

Finally to a point I raised earlier once I have a two star alignment - is the error in goto'ing a 3rd and final alignment star a reasonably good indication of polar misalignment (plus cone errors)?

Thanks all,

Matt

[Terry B]

Originally Posted by g__day

Terry - huge thanks for that - it's exactly what I needed to hear. As soon as the skies clear I'll try again.

BTW - I do have PoleAlignMax and the full DC3 pinpoint version - haven't experiment with them yet. I think I'll try a manual drift align - plugging my data into that little arc tangent function to see what drift corrections it advises then using Jupiter's diameter as my measuring post. I'll also give PEMPro another whirl now I know to use the SS2K in one star - polar aligned mode!

Terry out of interest - did you read that somewhere or were you able to deduce it from experience?

Finally to a point I raised earlier once I have a two star alignment - is the error in goto'ing a 3rd and final alignment star a reasonably good indication of polar misalignment (plus cone errors)?

Thanks all,

Matt

I use to own a SS2000 and remember it from then. It was discussed in the distant past on one of the vixen yahoo groups. I forget which one.

[g__day (Matthew)]

Well that was very interesting!

With a one point alignment in polar aligned mode, PEMPro consistently advised to move the mount West 5 arc minutes and lower its elevation 2 arc minutes. I've just finished doing this and of course clouds roll in - third night in a row.

At least I was able to let PEMPro run on both Az and Alt and seeing it suggest no more than 0.4 arc minute changes. I would have then liked to track a star for a while - so of course clouds have to pour in!

Two things have left me very surprised in all this:

1. How precise the SS2K's pointing and tracking correction algorithms are in a 3 star alignment - even when I'm near 8 arc minutes off the pole and doing long, long subs!

2. How stunning a lack in Vixen's documentation about how to set the SS2K if you want to drift align! This seems to have been a momumental oversight - especially as many folk on the yahoo SS2K group had no feedback on how to set up the SS2K for drift alignment!

When I have checked everything out I then have to re-build my Tpoint and MaxPoint models - which will raise the question about do I place the SS2K in polar aligned with a 1, 2 or 3 star model? Thoughts anyone? What I really hope to see is if I now try and do a 3 star alignment after the first two stars are locked in hopefully almost no misalignment error when I go to the 3rd star (meaning the SS2K isn't once again trying to operate in Alt Az mode)!

PS

Interesting to note - PEMPro still inverts the direction of the East / West corrections - but it does get the raise lower directions correct!

Cheers all,

Matt