Jump to content


Photo

Polar alignment 3 arc minutes acceptable?

  • Please log in to reply
47 replies to this topic

#26 orlyandico

orlyandico

    Fly Me to the Moon

  • *****
  • Posts: 5355
  • Joined: 10 Aug 2009
  • Loc: Singapore

Posted 23 June 2013 - 11:36 AM

I remember when I was renting on GRAS (now iTelescopes). The typical pointing accuracy, even with the T-Point modeling, the fixed observatory, etc. etc. etc. was about 1' to 2'. The reason I know this is because as part of the workflow, after a slew, the GRAS software would plate-solve and display how far off the pointing was. And these were Paramount ME's.

So I think Celestron saying 5' for the CGE Pro is probably the absolute best case, which would not be typically achievable by someone who just plunked it down in a field.

#27 orion69

orion69

    Viking 1

  • -----
  • Posts: 989
  • Joined: 09 May 2010
  • Loc: Croatia

Posted 23 June 2013 - 11:41 AM

Well, if you are satisfied with polar scope, that is perfectly OK.

Just to be clear, we are talking about Atlas or CGEM polar scope, not TAK.
Problem is not just polar scope alignment, also you can't be sure if drawings on polar scope reticle are correct... Also there is not enough magnification for polar scope to be precise enough.
You can achieve polar alignment with polar scope similar to ASPA but that involves certain amount of luck, ASPA in my experience is much more reliable.
Since nearly all my images are 30 min subs @ 765mm or 900mm difference is easily noticable.
I would really like to see 30 min subs image(s) that were done by mount (Atlas or CGEM) polar aligned with just polar scope.

#28 freestar8n

freestar8n

    Soyuz

  • *****
  • Posts: 3906
  • Joined: 12 Oct 2007

Posted 23 June 2013 - 11:42 AM

I thought modern mounts with good modeling were in the sub 1' range - perhaps 15" rmsd. For cge or cge-pro, a lot depends on how loaded it is. For an actual all-sky rmsd above perhaps 20 degrees altitude, I think 5' is not hard to do - and that's with 6 stars.

The main skill involved is in accurately centering the alignment and calibration stars, and finishing with up/right. And using two widely separate on one side for alignment - and four widely separate on the other side for calibration.

Frank

#29 orion69

orion69

    Viking 1

  • -----
  • Posts: 989
  • Joined: 09 May 2010
  • Loc: Croatia

Posted 23 June 2013 - 11:55 AM

The main skill involved is in accurately centering the alignment and calibration stars, and finishing with up/right. And using two widely separate on one side for alignment - and four widely separate on the other side for calibration.

Frank


Exactly, I think those are the main reasons why many people have problems with ASPA (and using wrong star for final polar alignment).
I use permanently mounted camera carefully perpendicular to scope and crosshairs in Maxim DL.

#30 neilson

neilson

    Viking 1

  • -----
  • Posts: 537
  • Joined: 22 Aug 2010

Posted 23 June 2013 - 11:58 AM

Hi,
I think the 5 arc minute pointing accuracy on the CGE Pro has probably improved since then. In recent firmware updates the pointing accuracy has been improved quit a bit. On my CGE, I usually check my ASPA with a drift alignment and rarely need to make any adjustments. Then I redo my 2+4 alignment and my goto's are dead center. But then my CGE wasn't made in China like the CGE Pro was. lol The Nexstar compensates for cone error automatically so I'm guessing its accuracy is pretty good on all their mounts. At least if you do the 2+4 alignment. And I never use the last star for my ASPA, it's rarely in the right place.

neilson

#31 orlyandico

orlyandico

    Fly Me to the Moon

  • *****
  • Posts: 5355
  • Joined: 10 Aug 2009
  • Loc: Singapore

Posted 23 June 2013 - 12:08 PM

The 5' for the CGE Pro is the current figure Celestron publishes on their web site - where they also claim +/- 3" PE.

#32 neilson

neilson

    Viking 1

  • -----
  • Posts: 537
  • Joined: 22 Aug 2010

Posted 23 June 2013 - 01:04 PM

Hi,
Yes you are correct. If it's still just 5 arc minutes then that could effect accuracy of the ASPA. Is it possible that after doing the 2+4 alignment the computer is able to take that in effect though.

neilson

#33 gdd

gdd

    Apollo

  • -----
  • Posts: 1410
  • Joined: 23 Nov 2005
  • Loc: N Seattle suburb, WA

Posted 23 June 2013 - 03:39 PM

Can we work the problem from the other direction, measure the polar alignment error from the tracking error?

Gale

#34 mikeschuster

mikeschuster

    Vostok 1

  • -----
  • Posts: 150
  • Joined: 25 Aug 2011
  • Loc: SF Bay area

Posted 23 June 2013 - 04:01 PM

Yes.
http://celestialwond...rErrorCalc.html

Rule of thumb: 1 arcsecond of drift for a declination 0 star near the meridian in 5 minutes equals about 1 arcminute of polar misalignment in azimuth.

#35 neptun2

neptun2

    Viking 1

  • -----
  • Posts: 805
  • Joined: 04 Mar 2007
  • Loc: Bulgaria

Posted 30 June 2013 - 08:40 PM

Well i use the new synscan polar alignment routine on my heq5 pro mount and am very happy with it. I don't know how correct is the reading of the synscan controller but after the polar align it is usually under 1 minute. I checked the result visually through the polar scope after that and the polar star is where it should be (i use polar finder program to see where it should be in the polar scope).

To sum it up - as other people said if you center the stars carefully and finish the movement with up and right keys the polar realign feature is working very good. I use the live view of my nikon d90 dslr which thanks to the grid allows me to center the stars with good precision.

#36 Cotts

Cotts

    Just Wondering

  • *****
  • Posts: 4778
  • Joined: 10 Oct 2005
  • Loc: Toronto, Ontario

Posted 30 June 2013 - 10:58 PM

If anyone is trying to achieve 5 arcmin polar accuracy or better and is not discussing the orthagonality of their mount (cone error) and the orthagonality of the telescope's attachment to the mount then they are ignoring significant sources of error.

Your polar scope is not nearly as accurate if the telescope doesn't point exactly the same direction as your mount's saddle. Look at your rings - is the felt the same thickness all around? Does each ring have the same thickness? Are the bolt holes centered to .01 inch tolerances or better? Your polar alignment will suffer also if your scope has cone error - the telescope/mount points to slightly different places in the sky when on exactly opposite sides of the mount.

Dave

#37 freestar8n

freestar8n

    Soyuz

  • *****
  • Posts: 3906
  • Joined: 12 Oct 2007

Posted 01 July 2013 - 01:36 AM

Cone and orthogonality don't affect the polar alignment. All that matters is that the single polar axis of rotation is pointed at the pole. The saddle could be 37 degrees off and the dec. axis bent 23 degrees - and as long as the polar axis is pointed right, the telescope will track well.

For all star polar alignment and other systems where terms like cone are modeled, you can use the goto accuracy as a gauge of how accurate the polar alignment will be - because it indicates how well that term is known and modeled.

So cone and dec. error do affect the way the mount behaves across the sky and they would need to be modeled - but ASPA does that - and an accurate polar scope doesn't need to know them in the first place.

But you would definitely want the polar scope well aligned in the polar axis - and 5' accuracy there isn't trivial to achieve.

Frank

#38 shawnhar

shawnhar

    Fly Me to the Moon

  • *****
  • Posts: 5357
  • Joined: 25 Jun 2010
  • Loc: Knoxville, TN

Posted 01 July 2013 - 09:39 AM

Well, if you are satisfied with polar scope, that is perfectly OK.

Just to be clear, we are talking about Atlas or CGEM polar scope, not TAK.
Problem is not just polar scope alignment, also you can't be sure if drawings on polar scope reticle are correct... Also there is not enough magnification for polar scope to be precise enough.
You can achieve polar alignment with polar scope similar to ASPA but that involves certain amount of luck, ASPA in my experience is much more reliable.
Since nearly all my images are 30 min subs @ 765mm or 900mm difference is easily noticable.
I would really like to see 30 min subs image(s) that were done by mount (Atlas or CGEM) polar aligned with just polar scope.

How about a 100% crop of a 10 minute sub at 2500mm using only the polar scope on an EQ6.
I normaly use the CCD Drift method but this session was only polar scope. It can be done.

Attached Files



#39 orion69

orion69

    Viking 1

  • -----
  • Posts: 989
  • Joined: 09 May 2010
  • Loc: Croatia

Posted 01 July 2013 - 11:14 AM

Of course it can be done, just not every time...
Since you are using CCD drift method I'm sure you agree with me. ;) :grin:

#40 shawnhar

shawnhar

    Fly Me to the Moon

  • *****
  • Posts: 5357
  • Joined: 25 Jun 2010
  • Loc: Knoxville, TN

Posted 01 July 2013 - 12:24 PM

Challenge accepted!
Now, would you please get rid of all the clouds and rain here so I can polar align each night using only the polar scope and report my findings. :cloudy: :bawling:

#41 freestar8n

freestar8n

    Soyuz

  • *****
  • Posts: 3906
  • Joined: 12 Oct 2007

Posted 01 July 2013 - 12:50 PM

Hi-

Your result may indeed be good - but it would help to know the actual fwhm in arc-seconds you obtained. Round is good - round and small, in arc-seconds, is even better.

I don't know what object that is, so I can't tell from context either. Is that a linear stretch?

Thanks,
Frank

#42 shawnhar

shawnhar

    Fly Me to the Moon

  • *****
  • Posts: 5357
  • Joined: 25 Jun 2010
  • Loc: Knoxville, TN

Posted 01 July 2013 - 02:50 PM

Sorry, that is M16, the Eagle. Used a Canon XS dslr.
One mild curve in PS and a little noise reduction.
Astrometry.net wouldn't plate solve it.

#43 orion69

orion69

    Viking 1

  • -----
  • Posts: 989
  • Joined: 09 May 2010
  • Loc: Croatia

Posted 01 July 2013 - 03:00 PM

Challenge accepted!
Now, would you please get rid of all the clouds and rain here so I can polar align each night using only the polar scope and report my findings. :cloudy: :bawling:


:)

Sorry I can't do that, same problem over here...
But it's easier for me since my new scope didn't arrive yet.
This year is really bad.

#44 Steve Mckean

Steve Mckean

    Lift Off

  • -----
  • Posts: 14
  • Joined: 22 Mar 2010
  • Loc: Bay Of Plenty, NZ

Posted 17 September 2013 - 10:28 PM

Hi Guys, sorry if this has been covered elsewhere but I was wondering why manually adjusting the mount directly to the NCP or SCP using plate solving and direct camera imaging would not work, the plate solving would let you know exactly which star you looking at through the main OTA so as long as you can keep the cross hairs in the center using manual adjustments would that not work ? (For polar alignment I mean).

Cheers

Steve

#45 frolinmod

frolinmod

    Surveyor 1

  • -----
  • Posts: 1804
  • Joined: 06 Aug 2010
  • Loc: Southern California

Posted 18 September 2013 - 02:07 AM

I remember when I was renting on GRAS (now iTelescopes). The typical pointing accuracy, even with the T-Point modeling, the fixed observatory, etc. etc. etc. was about 1' to 2'. The reason I know this is because as part of the workflow, after a slew, the GRAS software would plate-solve and display how far off the pointing was. And these were Paramount ME's.

That sounds like raw uncorrected pointing or a very poor pointing model. Maybe they weren't even enabling Tpoint pointing corrections at all. They're certainly not needed if the workflow includes plate solving and re-slewing as most do these days as a matter of course.

Also, they were probably using TheSky6 and Tpoint for Windows. TPoint for Windows was implemented using TPoint version 1 from Patrick Wallace. I'll bet their pointing model was fairly lousy because back then anything beyond a basic model had to be artfully hand crafted one term at a time, which was a PITA.

The TPoint Add On to TheSkyX is implemented using TPoint version 18. The new Tpoint supermodel feature automatically creates the pointing model for you, no artful hand crafting skills required. A new pointing model created by supermodel is likely to be far superior to any previously hand crafted one (especially when given a few hundred well distributed pointing samples to chew on).

#46 Phil Sherman

Phil Sherman

    Surveyor 1

  • -----
  • Posts: 1505
  • Joined: 07 Dec 2010
  • Loc: Cleveland, Ohio

Posted 19 September 2013 - 01:01 AM

Not yet! I have a guider I am going to add to the setup soon. It's a new pier and observatory I am now at the stage of aligning the mount before I start imaging.


If you're setting up the mount on a pier then you should be using your polarscope and/or any hand controller polar alignment routine for no more than a rough polar alignment. Once you've done this; you should be refining it using the drift method, preferably using your camera to measure the drift. Drift measurements are made along the celestial equator at the meridian (just W of the meridian) and close to either the E or W horizon. Close means around 30 degrees in altitude above the horizon. The meridian measurement is used to adjust azimuth while the horizon location is for altitude.

Set your tracking rate to sidereal, slew rate to 1x sidereal, exposure time to 70 seconds then start the exposure. Let the mount track for 5 seconds then slew E for 30 seconds, then slew W for 35 seconds. Look at the image. There will be stars (the 5 second tracked portion) with two trails forming a V shape. The size of the opening of the V at the star is an indication of how much drift occurred in 60 seconds. The relationship between the two trails will show you which way to adjust the mount. If the first adjustment makes the V wider, you moved the mount the wrong way. When the mount is polar aligned, the trails collapse into a line. At a minute per measurement, this technique is much much faster than using an eyepiece. It also makes it very easy to compare the current measurement with the last one.

If you're using an Atlas mount on a pier inside an observatory, you should be planning on running the mount with EQMOD. Everything to run the mount, including guiding is available as free downloads; EQMOD, ASCOM platform, CdC, PhD, and even AstroTortilla for plate solving. PhD can guide the Atlas using pulse guide, avoiding the need for an ST4 connection to the mount's guide port. This also allows webcams and any other camera that PhD can control to be used for guiding. You'll also want a motorized focuser that can be controlled from your computer. Orion has a relatively inexpensive geared DC motor unit that can be adapted to almost any focuser. Shoestring Astronomy makes a USB run controller for this motor that works with current Windows systems.

Phil

#47 orlyandico

orlyandico

    Fly Me to the Moon

  • *****
  • Posts: 5355
  • Joined: 10 Aug 2009
  • Loc: Singapore

Posted 19 September 2013 - 04:58 AM

the problem with the "V technique" is that it's not very accurate. the rule of thumb for declination drift is 1 arc-second in 5 minutes for 1 arc-minute of misalignment.

if your camera is 2"/pixel (a common size) in order to get a drift of 1 pixel (2") in 1 minute would require you to be out by 10 arc-minutes, which is a lot.

i have been doing this V-method for a long time and PEMPro polar alignment is bulletproof. Just leave it for 10-15 minutes for each of alt and az, and (assuming you didn't move the mount the wrong direction) you can get < 5' with one iteration of each Alt and Az.

also helps if you know how much (arc-minutes) the mount moves with every rotation of the adjuster bolts.

#48 Steve Mckean

Steve Mckean

    Lift Off

  • -----
  • Posts: 14
  • Joined: 22 Mar 2010
  • Loc: Bay Of Plenty, NZ

Posted 19 September 2013 - 04:13 PM

Thanks for that, ill work on it.

Steve






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics