Jump to content


Photo

CG5ASGT: 4min tracking at 100mm -> trails.

  • Please log in to reply
29 replies to this topic

#1 gezak22

gezak22

    Viking 1

  • *****
  • Posts: 897
  • Joined: 15 Aug 2004
  • Loc: On far side of moon. Send help.

Posted 13 September 2009 - 01:21 PM

Hey guys,

I apologize in advance for the long post.

As the title suggests, I use a CG5ASGT with a 100mm lens and I see trailing after only 4 minutes of exposing. Shouldn't I be able to go ~15min without seeing any trailing?

Here is my procedure. I use a 900mm FL refractor for aligning. The XSi with the 100mm lens is piggybacked on top of the refractor. Firmware is uptodate.

1. Level the mount. I have heard that the level in the mount can be off, so I bought an extra one. Where should I position it? Right on top of the internal bubble level? That's what I did for this run.

2. Sight Polaris through Polar scope hole. Make sure there is enough room to adjust to azimuth knobs. Make sure all the knobs are tight.

3. Turn on mount. Align markers on the axes. Enter all the info. I get the coordinates from Google Earth. Time could be off by +/- 5 minutes.

4. 2+4 Alignment. I use the stars the mount recommends. I heard that the shorter you see the 'Calibrating' text on the display, the better the alignment. After the first alignment star, it generally says 'Calibrating' for 7 seconds; after the last cal star, the text remains for 12 sec. I use a 12.5mm illuminated reticle and I always finish with the 'up' and 'right' buttons. Even the last cal star does not end up in the middle of the xhair. Should I Barlow the EP?

5. Go to a bright star near Meridian. 'Align' -> 'Polar Align' -> 'Align mount'. The mount now syncs on the star; eventually it asks me to adjust azimuth and altitude knobs to center star.

6. Turn mount off. I am not sure if this is the way to go. If you leave your mount on, how do you get it to do another 2+4 alignment?

7. Turn mount on. Align markers on axes. Update info (only time has changed).

8. 2+4 Alignment. Again, every star will need some adjusting to get to the center of the xhair and again I see the 'Calibrating' being displayed for ~12 sec on the last cal star.

9. At this point I start shooting.

As I said, I see trailing at 4 minute exposures. 3 min exposures look ok, but not perfect. I really though that the updated firmware and the external bubble level would solve this issue. Before the firmware update I was also limited to 3min exposures.

Any input would be greatly appreciated.

#2 rmollise

rmollise

    Hubble

  • *****
  • Posts: 15562
  • Joined: 06 Jul 2007

Posted 13 September 2009 - 03:39 PM

Hey guys,

I apologize in advance for the long post.

As the title suggests, I use a CG5ASGT with a 100mm lens and I see trailing after only 4 minutes of exposing. Shouldn't I be able to go ~15min without seeing any trailing?


I'd suggest a drift alignment as an experiment. Even if you're only at 100mm, less than dead-on alignment might reveal a little trailing with long unguided exposures.

Make sure you are really seeing trailing, too, not field curvature, etc.

Leveling the mount makes absolutely no difference, so don't waste time doing that.

#3 gezak22

gezak22

    Viking 1

  • *****
  • Posts: 897
  • Joined: 15 Aug 2004
  • Loc: On far side of moon. Send help.

Posted 13 September 2009 - 04:07 PM

Thanks for the input.

It is trailing. All the stars are streaks in the same direction. Even at f/5.

I thought leveling was crucial.

#4 rmollise

rmollise

    Hubble

  • *****
  • Posts: 15562
  • Joined: 06 Jul 2007

Posted 13 September 2009 - 04:11 PM

Thanks for the input.

It is trailing. All the stars are streaks in the same direction. Even at f/5.

I thought leveling was crucial.


Nope. Not at all.

#5 mclewis1

mclewis1

    Thread Killer

  • ****-
  • Posts: 10880
  • Joined: 25 Feb 2006
  • Loc: New Brunswick, Canada

Posted 13 September 2009 - 04:14 PM

It is trailing. All the stars are streaks in the same direction. Even at f/5.

But the real question now is which direction is it trailing in? E-W is generally a tracking issue, N-S an alignment issue.

#6 Charlie Hein

Charlie Hein

    Voyager 1

  • *****
  • Administrators
  • Posts: 12308
  • Joined: 02 Nov 2003
  • Loc: 26.06.08N, +80.23.08W

Posted 13 September 2009 - 04:25 PM

Rod's got some good advice - I'd definitely try it to see what's happening.

Here's a thought - it could still be misaligned but if you really think you did everything perfectly then I'd be checking to see if the tracking rate is correct - this issue could be as simple as having the mount set to Lunar or Solar rate.

#7 Blixx

Blixx

    Viking 1

  • *****
  • Posts: 531
  • Joined: 02 Dec 2007
  • Loc: CO. Aurora, USA

Posted 13 September 2009 - 04:29 PM

I have seen that if you want accurate goto's then you want the mount a level as possible. At least that is what I have observed with both the cg5 and the cgem.
Dan

#8 gezak22

gezak22

    Viking 1

  • *****
  • Posts: 897
  • Joined: 15 Aug 2004
  • Loc: On far side of moon. Send help.

Posted 13 September 2009 - 04:55 PM

Thanks, I'll look into these points.

In the procedure I listed, is 12.5mm at 900mm FL enough? Also, is step 6 ok?

#9 Jared

Jared

    Fly Me to the Moon

  • *****
  • Posts: 5049
  • Joined: 11 Oct 2005
  • Loc: Piedmont, California, U.S.

Posted 13 September 2009 - 05:36 PM

I have seen that if you want accurate goto's then you want the mount a level as possible. At least that is what I have observed with both the cg5 and the cgem.
Dan


That may be true for go-to's (depending on the pointing model built into the software), but it is definitely not a requirement for tracking. All you need for accurate tracking is proper polar alignment. Level/not level won't make a difference.

I agree with the previous recommendation--try a drift alignment and then repeat your experiment. Also, try to determine whether your drifting is in RA or Dec. That will help narrow down whether this is a tracking issue or an alignment issue.

#10 rmollise

rmollise

    Hubble

  • *****
  • Posts: 15562
  • Joined: 06 Jul 2007

Posted 13 September 2009 - 05:49 PM

I have seen that if you want accurate goto's then you want the mount a level as possible. At least that is what I have observed with both the cg5 and the cgem.
Dan


Haven't observed this at all. With the CG5, go-tos are more than accurate whether you are level or not or well polar aligned or not.

#11 gezak22

gezak22

    Viking 1

  • *****
  • Posts: 897
  • Joined: 15 Aug 2004
  • Loc: On far side of moon. Send help.

Posted 13 September 2009 - 06:38 PM

Alright, tracking rate is syderial and the streaks are almost exactly from NW to SE.

I was hoping I could avoid drift alignment, but I'll try it sometime this week.

Quick Edit: I should also point out that the 900mm department store scope wouldn't fit on the CG5 mounting bracket, so I had to drill an extra hole in the mounting bracket. This means that the scope is not exactly parallel with the mounting bracket. I don't think this is a problem, right?

#12 HaleBopper

HaleBopper

    Viking 1

  • -----
  • Posts: 768
  • Joined: 14 Jan 2008
  • Loc: Land of Ice and Snow

Posted 13 September 2009 - 06:57 PM

6. Turn mount off. I am not sure if this is the way to go. If you leave your mount on, how do you get it to do another 2+4 alignment?


You don't need to turn off the mount. keep hitting undo to scroll back to the start, and the HC will once again ask you to input your data, align index marks etc. This is because your previous alignment is lost when you go through the polar align routine.

#13 Patrick

Patrick

    Voyager 1

  • *****
  • Posts: 11323
  • Joined: 15 May 2003
  • Loc: Franklin, Ohio

Posted 13 September 2009 - 07:52 PM

You don't need to turn off the mount. keep hitting undo to scroll back to the start, and the HC will once again ask you to input your data, align index marks etc. This is because your previous alignment is lost when you go through the polar align routine.



You do not need to turn the mount off as noted above. At the end of the polar alignment routine, the goto alignment is not really 'lost', but it may be off a little if you had to move the mount very far to get polar aligned. Rather than start over, you can replace the alignment stars or recenter and sync on objects that may not be quite centered after doing a goto.



I use a CG5ASGT with a 100mm lens and I see trailing after only 4 minutes of exposing. Shouldn't I be able to go ~15min without seeing any trailing?



Without guiding, I think you should be happy getting 3 minute exposures at 100mm f/l. Have you tried generating a periodic error graph for your mount? The CG5's are typically around +/- 20 arcseconds PE (total 40 arcseconds), and sometimes more. More than likely it's the PE of the worm gear that's causing the trailing.

Depending on how dark your skies are, 2 to 3 minute exposures are actually quite good (assuming you're using a DSLR or CCD). You can then capture a bunch of images and stack them to reduce noise, and then stretch the image to bring out the detail. Rather than going for 15 minute exposures, try staying within the mount's limit and take multiple sub-frames.

Your other alternative is to autoguide.

Regards,

Patrick

#14 gezak22

gezak22

    Viking 1

  • *****
  • Posts: 897
  • Joined: 15 Aug 2004
  • Loc: On far side of moon. Send help.

Posted 13 September 2009 - 09:30 PM

Hey Patrick,

Are you sure 3 min @ 100mm is the max I can get out of this mount without guiding. The mount is way underloaded with my equipment. I understand the concept of stacking - I have been doing it for several years now.

#15 Panza

Panza

    Ranger 4

  • -----
  • Posts: 335
  • Joined: 26 Sep 2007
  • Loc: Lillehammer, Norway

Posted 14 September 2009 - 01:00 AM

It depends on what camera you are using too. arcseconds per pixel. If you're using a new-ish DSLR then you will likely have around 20 arcseconds pr pixel. That means your mounts PE won't make that much difference. I stay with most of the others here: Try drift alignment. I bet you'll be able to do at least 5 minutes. 10 should be doable at that scale.

#16 Patrick

Patrick

    Voyager 1

  • *****
  • Posts: 11323
  • Joined: 15 May 2003
  • Loc: Franklin, Ohio

Posted 14 September 2009 - 01:55 AM

Are you sure 3 min @ 100mm is the max I can get out of this mount without guiding. The mount is way underloaded with my equipment.



I'm not saying it is the max, but if PE is the problem, the weight is not going to make a difference. It is possible to overload a mount and hurt the performance, but underloading the mount still leaves the native PE of the mount in play. I'd suggest you measure your PE. If you have a webcam and a copy of K3CCDtools you can export Drift Explorer output files into a program called PEAS to generate a PE curve for your mount.

Once you know your PE range, then you can use Niel's Periodic Error Modeller to calculate wihat kinds of exposures will work for you.

The bottom line is that to improve performance, you first need to have the mount accurately polar aligned. You can do a hand controller PA routine, then check it by doing a drift alignment (shouldn't be too bad, since the HC PA routine should get it pretty close). Then, make sure you do not have any flexure in your setup that might cause movement. Finally, at that point, you are at the mercy of your mount's PE. The only way to not be at the mercy of the PE is to autoguide. If you're using a camera with small pixels, you might see trailing more quickly than if you're using a camera with larger pixels.

Patrick

#17 Tapio

Tapio

    Apollo

  • -----
  • Posts: 1327
  • Joined: 24 Sep 2006
  • Loc: Tampere, Finland

Posted 14 September 2009 - 02:46 AM

I think 3 minutes @100mm with CG5ASGT should be quite easy task.
Here's one example with 3 min exposures @70mm and with EQ2 mount (only crudely polar aligned).
http://www.ursa.fi/c...yway_140809.htm

Your alignment procedure sounds ok to me but I'm sure if Celestron's PA procedure is acurate enough for photographic purposes.
As others have said you should try drift alignment.

#18 Patrick

Patrick

    Voyager 1

  • *****
  • Posts: 11323
  • Joined: 15 May 2003
  • Loc: Franklin, Ohio

Posted 14 September 2009 - 11:36 AM

It depends on what camera you are using too. arcseconds per pixel. If you're using a new-ish DSLR then you will likely have around 20 arcseconds pr pixel. That means your mounts PE won't make that much difference. I stay with most of the others here: Try drift alignment. I bet you'll be able to do at least 5 minutes. 10 should be doable at that scale.



Yes, it depends on the camera and focal length, but also on the PE. For a Canon XSi, at 100mm f/l each pixel sees about 10 arc seconds. (See this Calculator for DSLR Astrophotography and insert your own camera's parameters). I think the worm gear cycle on a CG5 is around 8 minutes. If the PE is +/- 20 arc seconds, then in 8 minutes there will be a total of 40 arc seconds of variation from high to low. At 4 minutes there will be about 20 arc seconds of PE on average. If the camera sees 10 arc seconds/pixel, even 4 minutes will start to show trailing.

That's why it's important to know what the mount's PE is if you're doing unguided imaging. But, all that aside, your images are really already telling you something important...3 minutes are good, 4 minutes aren't.

I'll go back to what I said earlier...it's unclear to me why you want to do subs longer than 3 or 4 minutes at 100mm f/l. In my area, I'm limited to about 90 seconds before skyglow overwhelms my images. If you're in dark skies, you can go longer, but then you can capture more data with shorter exposures anyway. I'd rather take a lot of shorter exposures and reduce noise, than have a couple of really long exposures with a bunch of noise. That's my 2 cents. :smirk:

Regards,

Patrick

#19 gezak22

gezak22

    Viking 1

  • *****
  • Posts: 897
  • Joined: 15 Aug 2004
  • Loc: On far side of moon. Send help.

Posted 14 September 2009 - 08:32 PM

The real reason behind trying to do 8min exposures was to make a HDR image. I guess I can still give that a try as long as I just play with the ISO instead of exposure time.

The other reason was that the bigger the drift, the smaller the "effective" imaging area. Once you stack the images, the edges look underexposed. The bigger the drift, the bigger the area that looks underexposed (because the overlap between 1st and last sub is smaller).

Another reason was that I wanted to see if longer exposures would bring out more details in really faint nebulae (or dark nebulae for that matter). One major target for winter is the Taurus molecular cloud, which might benefit from the longer subs. The cold temperatures should keep noise down.

Another point. If I hit 'Undo' often enough, the CG5 only takes me back to a screen that reads 'Advanced GT', I don't see an option to enter all the info and do another 2+4 alignment. Not that it matters too much - turning the mount off and on again may take 1 second longer than the other way.

Edit: This now brings me to an earlier question that was left unanswered. I will try drift alignment. For drift alignment it is advantageous to level the mount to reduce the number of iteration. Suppose I don't trust the bubble level of the CG5, where on the mount should I place the external level?

Thanks.

#20 Patrick

Patrick

    Voyager 1

  • *****
  • Posts: 11323
  • Joined: 15 May 2003
  • Loc: Franklin, Ohio

Posted 14 September 2009 - 09:59 PM

For drift alignment it is advantageous to level the mount to reduce the number of iteration. Suppose I don't trust the bubble level of the CG5, where on the mount should I place the external level?



Yes, that's true. When I'm on uneven ground, I'll first throw a level on top of the tripod before putting the mount on, making sure the feet are well seated. If I put the mount on without leveling first, I'll put the level on the eyepiece tray/leg spreader. The problem with both of those scenario's is that when I set up, it's usually not dark yet so I may have to move the mount to get it rough aligned to Polaris, which can throw the level off...which means leveling it again.

The other reason was that the bigger the drift, the smaller the "effective" imaging area. Once you stack the images, the edges look underexposed. The bigger the drift, the bigger the area that looks underexposed (because the overlap between 1st and last sub is smaller).



Are you talking about light drop off at the edges of the image? If so, that's probably vignetting due to not having 100% field illumination on the XSi sensor. Are you taking flats? That usually helps with vignetting. I don't understand what you mean by 'drift', unless you're referring to the movement of the stars from frame to frame. That kind of drift is actually a result of polar misalignment. The stars may not be trailing in the image because the exposure was short enough not to capture it, but over time the star field moves from frame to frame. That brings you back to the need for a good polar alignment. :smirk:

Another point. If I hit 'Undo' often enough, the CG5 only takes me back to a screen that reads 'Advanced GT', I don't see an option to enter all the info and do another 2+4 alignment. Not that it matters too much - turning the mount off and on again may take 1 second longer than the other way.



The hand controller has a hierarchical structure with the 'Advanced GT' screen at the top. If you hit the 'Align' button to the left of the 'Enter' key, you can replace both Alignment and Calibration stars. You will goto the star first (Named Stars...under List), then hit the Align button, scroll down to either Alignment Star or Calibration Star and replace the old star with the new star (or overwrite the location of the old star with the same star at the new location.

Of course you can do what you suggested, which is turn the switch off, manually move the scope to home position and redo the 2-4 alignment, or, you can just goto the new star and replace the old one with the new one.

If you need a manual, you can download one at Celestron's website or at the Nexstar Resource site.

Patrick

#21 gezak22

gezak22

    Viking 1

  • *****
  • Posts: 897
  • Joined: 15 Aug 2004
  • Loc: On far side of moon. Send help.

Posted 14 September 2009 - 10:40 PM

Are you talking about light drop off at the edges of the image? If so, that's probably vignetting due to not having 100% field illumination on the XSi sensor. Are you taking flats? That usually helps with vignetting. I don't understand what you mean by 'drift', unless you're referring to the movement of the stars from frame to frame. That kind of drift is actually a result of polar misalignment. The stars may not be trailing in the image because the exposure was short enough not to capture it, but over time the star field moves from frame to frame. That brings you back to the need for a good polar alignment. :smirk:


I am not talking about light drop off at the edge of the sub but about the movement of the stars from frame to frame. I would like this to be as small as possible, so I need a better polar alignment routine.

If I replace one alignment star as you suggested, will it automatically ask me to replace the other one, and the 4 calibration stars? Sorry for asking this, I am too lazy to haul the 12V battery to the mount to actually try it out.

#22 amateur

amateur

    Messenger

  • *****
  • Posts: 413
  • Joined: 15 Jun 2006
  • Loc: The Netherlands

Posted 15 September 2009 - 05:21 AM

What is the periodic error of your mount? If your polar scope works OK, than periodic error is the dominant source of trailing.

You can simulate that using this simulator..
The snapshot contains some values that are typical for your setup:

- 50 arcsec PE amplitude (could be worse)
- 5 minute exposure
- polar alignment accurate within 10 arc minutes to the East (already a worst case scenario if you polar scope works OK)
- 100mm lense
- typical pixel size of 5 microns:

Attached Files



#23 amateur

amateur

    Messenger

  • *****
  • Posts: 413
  • Joined: 15 Jun 2006
  • Loc: The Netherlands

Posted 15 September 2009 - 05:24 AM

In practice I would expect that the width of the star trails is 15 arc seconds at least (likely much more). So effectively the PE trailing will look like a line.

Attached Files



#24 amateur

amateur

    Messenger

  • *****
  • Posts: 413
  • Joined: 15 Jun 2006
  • Loc: The Netherlands

Posted 15 September 2009 - 05:27 AM

And if your worm start at another angle (phase of the PE period), it looks even more like trailing:

Anyway it is very unlikely that the trail will have a SE direction (45 degrees to the horizontal) unless your polar alignment is a big mess, or the rate of your drive is off.

Attached Files



#25 Patrick

Patrick

    Voyager 1

  • *****
  • Posts: 11323
  • Joined: 15 May 2003
  • Loc: Franklin, Ohio

Posted 15 September 2009 - 05:28 AM

If I replace one alignment star as you suggested, will it automatically ask me to replace the other one, and the 4 calibration stars?



No, you have to replace each star one at a time by going to the replacement star and doing the align or calibrate star routine.

Patrick






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics