Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

cone error and 3 star alignment

  • Please log in to reply
14 replies to this topic

#1 Tom3

Tom3

    Explorer 1

  • -----
  • topic starter
  • Posts: 62
  • Joined: 23 Jan 2018
  • Loc: Scottsdale, AZ

Posted 19 September 2019 - 12:11 AM

How much does 3-star alignment correct for cone error?

 

Thanks for any info or observations.

 

Tom



#2 J A VOLK

J A VOLK

    Apollo

  • *****
  • Posts: 1074
  • Joined: 24 May 2006
  • Loc: So. Cal.

Posted 19 September 2019 - 12:48 AM

It should correct pretty well if the cone error is not severe. My CEM120/C14 shows a cone error of about .5 degrees and points extremely well with 3-star alignment, much better than 1 or 2 star. If the cone error changes with position due to flexibility or free-play then you certainly won't get a good result. I guess this could be tested by doing several alignments with different sets of stars to see if the results are similar. I don't know if other brands give a result for cone error like Ioptron.

#3 sg6

sg6

    Fly Me to the Moon

  • *****
  • Posts: 5910
  • Joined: 14 Feb 2010
  • Loc: Norfolk, UK.

Posted 19 September 2019 - 02:13 AM

I am not sure is the only answer I can give.

In general 2 star alignment seems as good as 3 star. Also not sure that 3 star alignment is always "3" star either.

 

To me it is the movement/difference between the stars that is the important bit and in cases the first star is simply a start position. You have to have a known/defined start position to determine the corrections needed to get star 2 centered. Where or what star 1 is is then irrelevant as it in itself does nothing.

 

In effect you start at some initial position (happens to be a star) Move to first measured star, determine corrections, move to second measured star to refine corrections. I would consider that as 2 star, whereas it is given as 3 star.

 

The system is moving S1 to S2, determining corrections the S2 to S23 and refining corrections. There are 2 movements, but you need 3 points to move between. Scopes happen to use stars (well usually).

 

The Meade 2 star alignment shows this as their "first" star is in effect a virtual star on the horizon at due North. Theirs is therefore Virtual star, first star, second star. 3 stars ???



#4 astronz59

astronz59

    Viking 1

  • -----
  • Posts: 953
  • Joined: 15 Sep 2010
  • Loc: Auckland, New Zealand

Posted 19 September 2019 - 06:04 AM

Not sure if I'm understanding this correctly, but I thought 'cone error' refers to a situation where the optical axis of the telescope is not parallel to the declination axis. In which case, cone error would affect tracking, not GOTO accuracy. Cone error is easily checked and corrected:

  1. Select a star near the equator and the meridian.
  2. Slew the telescope to the East side of the mount and center the star. Note the time and the RA reading and the time.
  3. Slew to the West side and repeat. If the difference between the two RA readings,differs from the time interval between readings, the telescope is not orthogonal to the Dec axis.
  4. If the difference is greater than the time interval, then, with the telescope E, the objective must be moved away from the mount and vice versa.

  • Kegan128 and Der_Pit like this

#5 EFT

EFT

    Vendor - Deep Space Products

  • *****
  • Vendors
  • Posts: 5250
  • Joined: 07 May 2007
  • Loc: Phoenix, AZ

Posted 19 September 2019 - 11:04 AM

If you are asking to see if you can get away without doing a three-star alignment on a new setup (i.e., not permanently mounted), then just run the three-star.  A lot of people ask if they can use fewer stars to align with (e.g., can I do less than the 2+4 on a Celestron mount?) and the answer should be no.  There is a reason to add additional stars to an alignment model when they are available and if you want the best performance from your mount (which you should want), then you use the best alignment available.  What's the hurry?  Ask yourself if you want good goto's and decent tracking (even for visual use) and if the answer if yes, then spend a small amount more time when you start up.  The amount of time necessary to complete a full alignment is not that long as is well spent.  Relax and take your time.  This is a patient, not rushed hobby.  If a two star alignment does just as well as a three-star alignment, that just means you don't have very much cone error to start with.  It does not mean that a two-star alignment is just as good as a three-star alignment in any general sense.

 

Now if you are asking because you think you have a lot of cone error and it is not being accounted for, then I not sure just how you quantify that.  Adding a star to compensate for cone error does just that.  While there may be some theoretical limit, it doesn't matter because you would never be so far off because it would be obvious to you to start with.  Something you have to keep in mind though is that, even if you account for cone error, the measurement of and compensation for it will still be limited by the quality of the mount and the quality of the centering of stars (e.g., using a high-powered reticle eyepiece as opposed to a defocused star for centering).


  • Cyrano, jdupton and SkipW like this

#6 Tom3

Tom3

    Explorer 1

  • -----
  • topic starter
  • Posts: 62
  • Joined: 23 Jan 2018
  • Loc: Scottsdale, AZ

Posted 20 September 2019 - 03:03 PM

Thanks for the helpful replies.

 

Tom



#7 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 8823
  • Joined: 12 Oct 2007

Posted 20 September 2019 - 05:41 PM

I think an important thing to realize is that 3 stars may work well one time but not another.  People fixate on cone error as a key problem, but another issue is the declination offset.  That may seem like a trivial issue that just shifts the stars north or south a bit - but it is similar to cone error in terms of pointing and needs to be determined accurately for a good model of the mount.

 

So in general you have polar alignment error in alt and az - and you have cone and dec. offset - along with an error in time or ra offset.  That is 5 unknowns right there and each star provides two measurements.  As a result, three stars is the bare minimum to determine all this.  The measurements all have some error - so with only three stars the model could end up far off.  And there are additional mount terms that play a role and may act like cone.

 

Unless the mount is very high end and has little cone error or dec. offset - I think 5 stars is about the minimum for good all sky pointing on every try.  But if you are lucky, three stars will do well sometimes.  And note that both cone and dec. offset involve the OTA and how well the optic axis is aligned with the tube - so it goes beyond the quality of the mount and depends on the optics also.

 

Another thing is that with Celestron 2+4 alignment, some people will stop adding stars once a new star lands in the center of the field.  I would go ahead and add that star to the model and not just say 'ok - the model is good enough.'  It may just happen to get the one star centered - but it doesn't mean the next star will also be centered.  And adding the star to the model will help refine it.

 

Frank


  • SkipW likes this

#8 Tom3

Tom3

    Explorer 1

  • -----
  • topic starter
  • Posts: 62
  • Joined: 23 Jan 2018
  • Loc: Scottsdale, AZ

Posted 20 September 2019 - 07:11 PM

Frank,

Can you explain declination offset in this context?  I am using a CEM25p which has no setting circles if that has anything to do with this.

 

Thanks,

Tom



#9 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 8823
  • Joined: 12 Oct 2007

Posted 20 September 2019 - 07:33 PM

Frank,

Can you explain declination offset in this context?  I am using a CEM25p which has no setting circles if that has anything to do with this.

 

Thanks,

Tom

The dec. offset is almost never mentioned as important for goto accuracy but it's actually critical and hard to determine.  Each axis has some angle being reported but you don't know how well they are referenced to 0 or 90 degrees - and you have to figure it out by measuring stars in the same way you have to figure out cone.

 

Cone will cause motion in the dec. axis to make a big, wide cone in the sky instead of a great circle.  So it affects all motions and all angle measurements - and it isn't just something that you notice when you do a meridian flip.  Similarly if you think the dec. axis is exactly 0 degrees and is pointing at the equator - but instead it's a bit off - motions in RA won't be a great circle either.

 

The RA offset is less critical because it really is just an offset in the sky and is equivalent to a slight time error.  But the dec. axis offset affects how far the mount moves in the sky when you move in RA - and you need to figure out what the value is in order to compensate for it and have accurate goto.

 

It doesn't matter if the mount has setting circles.  Some mounts have homing switches that allow a repeatable initial setup very near where the dec axis is pointing to the pole.  But internally the goto software will need to figure out the true dec. offset as stars are added.

 

So there are at least 5 terms needed for good goto - and you need a bunch of stars to nail them down with confidence.

 

Frank


  • SkipW likes this

#10 SkipW

SkipW

    Messenger

  • -----
  • Posts: 473
  • Joined: 03 Feb 2011
  • Loc: Oklahoma, USA

Posted 20 September 2019 - 10:43 PM

Those are great explanations, freestar8n!

 

What you call declination offset, I had heard called "declination index error", or just "index error", and yes, it only really matters in dec (or elevation angle in an Az-El mount), since a similar offset in RA (or Az) gets factored out in a two-star alignment, even if that seems counterintuitive.



#11 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 8823
  • Joined: 12 Oct 2007

Posted 21 September 2019 - 01:13 AM

Those are great explanations, freestar8n!

 

What you call declination offset, I had heard called "declination index error", or just "index error", and yes, it only really matters in dec (or elevation angle in an Az-El mount), since a similar offset in RA (or Az) gets factored out in a two-star alignment, even if that seems counterintuitive.

Thanks!  Yes I think there are different names for it.  For Celestron mounts it is shown in the mount settings menu as "dec index" and it is right next to Cone.  

 

There are several tools and methods I have seen on the web intended for measuring cone error in a mount - but they won't work well because they assume everything else about the mount is perfect - such as the angle between the axes being exactly 90 degrees when it isn't.  But measuring a decent number of stars around the sky should find it pretty well.  Three is the bare minimum and will probably have hit and miss results.

 

Frank



#12 Der_Pit

Der_Pit

    Viking 1

  • -----
  • Posts: 843
  • Joined: 07 Jul 2018
  • Loc: La Palma

Posted 21 September 2019 - 12:15 PM

One (hopefully related, and not too stupid) question here:  Does it matter how/where the 3-star alignment is done?  Where are the results stored?  Is it in the mount firmware, in the hand controller, or the external control software?

In the simple case, could I do a 3-star with the hand controller of my CEM60, and then disconnect it, use external software via the serial port and get correct pointing? 



#13 ThatsMyCoffee

ThatsMyCoffee

    Ranger 4

  • -----
  • Posts: 398
  • Joined: 15 Dec 2017
  • Loc: Czech Republic

Posted 21 September 2019 - 03:32 PM

Where are the results stored?  Is it in the mount firmware, in the hand controller, or the external control software?

In the simple case, could I do a 3-star with the hand controller of my CEM60, and then disconnect it, use external software via the serial port and get correct pointing?

Firmware is "read-only".  Definitely not there.  It contains the program that the mount uses to operate.  But it does not include temporary data.  Just the program.  It's not altered until a new version is written in it's place.

 

Whatever you used to build the model, will determine where the model is 'held'.  If you did it on the hand controller, it'll be stored in the mount.  If you did it via external software, that's where it will be.

 

In the cases mentioned previously, I think it was assumed in the hand controller.



#14 Der_Pit

Der_Pit

    Viking 1

  • -----
  • Posts: 843
  • Joined: 07 Jul 2018
  • Loc: La Palma

Posted 21 September 2019 - 04:47 PM



Firmware is "read-only".  Definitely not there.  It contains the program that the mount uses to operate.  But it does not include temporary data.  Just the program.  It's not altered until a new version is written in it's place.

Well, with 'firmware' I referred to the mounts' control boards.   And they do store some stuff that can be changed, like park position or zero position, even over power downs.  And also variables like tracking speed, guide speed etc. are kept on the board while it is running.



#15 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 8823
  • Joined: 12 Oct 2007

Posted 21 September 2019 - 05:38 PM

There are different ways to implement a 3-star alignment so it will depend on the software in the controller.  But normally you would pick 2 stars on one side of the meridian and a third on the other side - and they should be widely separated in dec. and ra.

 

For Celestron mounts there is a distinction between 'alignment' stars and 'calibration' stars.  Alignment stars are on one side and cal stars are on the other.

 

Once you do a full 2+4 alignment (or anything 2+1 or more) the mount has an idea of cone and dec. index and those numbers are stored in eeprom in the handcontrol.  So if you set up again with the same ota but a different polar alignment, you can just do a two-star align and it will reuse the previous calibration terms that were stored.

 

But other mounts doing 3-star alignments may not store the info and you need to redo it every time.

 

The better you know the calibration terms, the better you will know how far of your polar alignment is.  With only three stars I wouldn't put too much confidence in how far off it thinks your alignment is.  That is one reason people find a big discrepancy between what the mount reports based on a few stars, vs. other methods these days that measure the polar alignment more directly.  The methods that just involve rotating the ra axis and taking images around the pole don't need to know cone or anything like that - and they can get a more direct measure of the polar alignment error.  But a Celestron mount with 2+4 should also do pretty well.

 

Frank


  • Der_Pit likes this


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics