Jump to content


Photo

Dec Cogging on CG-5?

  • Please log in to reply
22 replies to this topic

#1 Madratter

Madratter

    Fly Me to the Moon

  • *****
  • Posts: 6842
  • Joined: 14 Jan 2013

Posted 06 May 2013 - 02:19 PM

I have a CG-5 and auto-guiding in Declination has been a real exercise in frustration. PHD will send guiding pulses but nothing seems to happen. Then eventually, the declination axis gets far enough off, and I get a big correction back. In other words, I get the same kind of sawtooth pattern people have been reporting on the CGEM.

At first, I thought this was because of weight balancing issues in declination. However, I now have my scope setup so that is licked.

The only way I have found to get decent long exposure photographs with this setup (5m) is to make sure I have very accurate polar alignment, i.e. I need to drift align first.

Any ideas? This is costing me WAY to many frames doing my astrophotography.

Also, and possibly related, when calibrating with PHD, I get the movement along the N axis, but it fails to return very far at all doing the S calibration. Again, I thought that was being caused by weight balance issues, but that does not appear to be the case. I have tried up backlash settings to around 30, but that does not change things.

I would love to chuck this mount and get an Atlas, but for the usual reason, that is currently out of the question.

There is a lot I love about this mount (great gotos is one), but auto-guiding in declination is driving me bug-eyed.

#2 Whichwayisnorth

Whichwayisnorth

    Surveyor 1

  • *****
  • Posts: 1525
  • Joined: 04 Jul 2011
  • Loc: Southern California

Posted 06 May 2013 - 02:23 PM

Have you taken the cover of the gearbox so you could watch what happens? Can you set the controller to guide rate and manually spin the motor? My thought is that if the gear train including but not limited to the worm and ring are binding or are too tight that may cause the issue. Perhaps cleaning out all the factory grease and replacing it with superlube and adjusting the gear meshing would benefit here.

This is all specualtion on my part. ymmv.

#3 rmollise

rmollise

    Hubble

  • *****
  • Posts: 15704
  • Joined: 06 Jul 2007

Posted 06 May 2013 - 02:24 PM

This isn't cogging, but backlash. The usual cure is to guide _only_ in one direction in dec. FWIW, the calibration of the CG5 always took a while with PHD, but I never had any problem guiding.

#4 gmartin02

gmartin02

    Viking 1

  • *****
  • Posts: 866
  • Joined: 11 Apr 2005
  • Loc: Santa Clarita, CA

Posted 06 May 2013 - 03:32 PM

Also, and possibly related, when calibrating with PHD, I get the movement along the N axis, but it fails to return very far at all doing the S calibration.

When I use PHD, the South calibration never moves back to the center of the cross hairs/box, but as soon as it finishes, PHD will "jump" to the middle, and guides correctly in Dec. I always thought this is some kind of screen refresh issue with PHD in South calibration only - the other 3 directions always show movement on the screen.

#5 Madratter

Madratter

    Fly Me to the Moon

  • *****
  • Posts: 6842
  • Joined: 14 Jan 2013

Posted 06 May 2013 - 03:40 PM

Thanks guys. I will need to take the motor housing off and check what is happening.

Rod, it makes sense to me the problem with PHD is backlash. But I don't understand why I would get the sawtooth pattern with backlash. It is guiding consistently in one direction, and I still get that sawtooth pattern.

At least with the the CGEM, that is blamed on cogging, which is why I asked.

Anyway, all that probably means is I don't understand what is going on properly.

#6 tjugo

tjugo

    Viking 1

  • -----
  • Posts: 995
  • Joined: 06 Nov 2007

Posted 06 May 2013 - 03:45 PM

This is not backlash, it it were backlash it will take a while to start moving but the movement will be smooth.

The problem is stiction and has to do with the lack of bearings in the DEC axis.

As Mr. Rod said, guiding only in one direction will help. MaximDL have an anti-stiction setting in the guiding parameters, this is very useful cause the software will figure out for you the DEC drift direction and never reverse the DEC corrections.

You can always tune your mount to make things smoother, but with practice, software and frustration you can overcome this issue.

Cheers,

Jose

#7 Madratter

Madratter

    Fly Me to the Moon

  • *****
  • Posts: 6842
  • Joined: 14 Jan 2013

Posted 06 May 2013 - 03:56 PM

Would stiction show up as a very regular sawtooth though? I can understand how the guiding would be very jerky, but I don't see how you would get that regular sawtooth.

#8 tjugo

tjugo

    Viking 1

  • -----
  • Posts: 995
  • Joined: 06 Nov 2007

Posted 06 May 2013 - 04:04 PM

It does...

Because of the lack of bearings, when you try to move the mount in DEC it will require several pulses from the guider, then when the stiction breaks the mount will "jump", and the guider will start sending commands in the opposite direction and the same thing will happen, after several commands the stiction will break and the mount will jump again.

Cheers,

Jose

#9 rmollise

rmollise

    Hubble

  • *****
  • Posts: 15704
  • Joined: 06 Jul 2007

Posted 06 May 2013 - 04:10 PM

I'm also thinking balance will help--as in NOT perfect declination balance. I always went a little tail heavy. ;)



#10 Madratter

Madratter

    Fly Me to the Moon

  • *****
  • Posts: 6842
  • Joined: 14 Jan 2013

Posted 06 May 2013 - 04:29 PM

Rod: I have tried fooling with the balance some, now that I can (I use the ADM counterweight set), without any resolution so far. But I probably have not tried every possible combination yet (declination going N or S, tail or front heavy).

Jose: I'm definitely wishing I had taken a screen shot of what is happening. The sawtooth is all on one side of the axis. It isn't guiding in opposite directions. It will drift in a direction getting larger and larger corrections sent. Once the corrections get large enough (somewhere around 500ms, then it will finally correct back to the middle) where it is supposed to be. Then the process starts all over again. I need to find my PHD log files and software to display it so I can show what is happening, but it looks exactly like what someone posted the CGEM declination cogging looks like, which is why I asked.

#11 Falcon-

Falcon-

    Aurora

  • *****
  • Posts: 4869
  • Joined: 11 Sep 2009
  • Loc: Gambier Island, BC, Canada

Posted 06 May 2013 - 04:36 PM

Yes absolutely sticktion can and does cause that sawtooth - it certainly did with mine! I took mine apart and re-greased it and it reduced the problem quite a bit. The guiding in one directly only trick also helps a lot.

Ultimately I had my best CG5 guiding results when I had several imaging nights in a row and left my equipment set up each day. Each night I slightly improved the polar alignment and by the last night the polar alignment was so good I was able to turn DEC guiding off entirely and only guide in RA!

(The lesson of course being to learn from my mistake and spend moer time on the FIRST night doing drift align or multiple iterations of the ASPA function)

Oh, and a hypertune service like DeepSpacePoducts may help as well if you have spare $$ as he not only does re-grease but also adjustments and polishing of bearing surfaces.

#12 nganga

nganga

    Vostok 1

  • -----
  • Posts: 116
  • Joined: 16 Mar 2008

Posted 06 May 2013 - 05:20 PM

Yes absolutely sticktion can and does cause that sawtooth - it certainly did with mine! I took mine apart and re-greased it and it reduced the problem quite a bit. The guiding in one directly only trick also helps a lot.

Ultimately I had my best CG5 guiding results when I had several imaging nights in a row and left my equipment set up each day. Each night I slightly improved the polar alignment and by the last night the polar alignment was so good I was able to turn DEC guiding off entirely and only guide in RA!

(The lesson of course being to learn from my mistake and spend moer time on the FIRST night doing drift align or multiple iterations of the ASPA function)

Oh, and a hypertune service like DeepSpacePoducts may help as well if you have spare $$ as he not only does re-grease but also adjustments and polishing of bearing surfaces.


I don't doubt that stiction could cause this, but I do have a strong suspicion that cogging occurs in the CG5 as well. For one thing, IIRC, the CGEM and CG5 have the same motor. Also, I did have my CG5 hypertuned by Ed Thomas. This did not cure the declination guiding problem. Longer and longer pulses would be delivered, then there would be a large correction. Some nights would be worse (or better) than others. Tossing any number of subs has been quite routine because of this.

I have read that when PHD does not return to the original starting point after calibration that is a sign of backlash, but backlash should have been taken up after PHD has been allowed to "settle down", according to the PHD Guiding guide.

Clem

#13 rmollise

rmollise

    Hubble

  • *****
  • Posts: 15704
  • Joined: 06 Jul 2007

Posted 06 May 2013 - 05:29 PM

[
I don't doubt that stiction could cause this, but I do have a strong suspicion that cogging occurs in the CG5 as well.


I suppose it's possible that Celestron might have changed something over the years, but I've never heard of this with a CG5, and mine (2005) has certainly never exhibited such behavior. I always imaged at modest focal lengths though--1300mm max, so I might not have noticed. My experience guiding the mount was that calibration generally took a while, but that guiding was pretty dependable, including with my ST2000 at 1300mm.

#14 Peter in Reno

Peter in Reno

    Fly Me to the Moon

  • *****
  • Posts: 6079
  • Joined: 15 Jul 2008
  • Loc: Reno, NV

Posted 06 May 2013 - 05:56 PM

Excessive backlash can also show sawtooth pattern. My previous CPC0800 had severe Declination backlash so I increased PHD Max Dec Duration to 1500msec and it helped a lot. Try increasing PHD's Max Dec Duration to 1500 - 2000msec for your mount. This feature is designed to handle excessive backlash/stiction.

Peter

#15 Madratter

Madratter

    Fly Me to the Moon

  • *****
  • Posts: 6842
  • Joined: 14 Jan 2013

Posted 06 May 2013 - 07:31 PM

I image at 1260mm so I'm not at real long focal lengths but not real short either. My mount was presumably made last year since I acquired it toward the end of the year.

In terms of backlash and setting Max Dec Duration, I don't think that will help me much since it does correct before it gets to those kind of huge values.

I'm wondering if maybe the nut on the end of the declination axis is just too tight. I might play with that as well as looking at motor to determine what if anything is going on with it.



#16 Madratter

Madratter

    Fly Me to the Moon

  • *****
  • Posts: 6842
  • Joined: 14 Jan 2013

Posted 06 May 2013 - 07:59 PM

I just pulled the motor housing and watched (and listened) to what is happening.

Something smells very fishy. (That covers 3 of the senses).

First, there is a huge amount of backlash. When I press a direction opposite of what I had, my backlash setting kicks in. (I believe I currently have this set to 20 both positive and negative).

Then the motor sits and does NOTHING for at least a second and I think more like a couple of seconds.

Then it starts to whir and move.

So the question is why does it do NOTHING for at least a second.

It is no wonder my guiding in declination is horrible.

EDIT: That reminds me of something else I had completely forgotten about until now. Originally when I was setting up PHD, I was getting no calibration on the Declination Axis. It was only once I really upped the calibration step size (to 1500) that it would eventually calibrate in declination. It might calibrate at a lower value. I don't know. But it would not calibrate in declination at the default.

#17 Madratter

Madratter

    Fly Me to the Moon

  • *****
  • Posts: 6842
  • Joined: 14 Jan 2013

Posted 06 May 2013 - 08:52 PM

Well, I learned a whole lot by taking the motor cover off and watching what is happening.

It turns out that if the declination backlash settings are high, at least on my mount, the motor will make an initial big leap, and then do nothing for a while. However, at smaller backlash settings this does not occur. I have moved my backlash settings back down to 5.

I don't know if this will help with my sawtooth behavior when guiding or not.

I suspect not, unless in fact it is switching directions. I don't think that is the case. But either way, this needed fixing.

Watching the motor, I do not think I am seeing cogging, at least as I have seen it described.

#18 Raginar

Raginar

    Fly Me to the Moon

  • *****
  • Posts: 6138
  • Joined: 19 Oct 2010
  • Loc: Rapid CIty, SD

Posted 06 May 2013 - 11:27 PM

I think I remember something with my CGEM about turning backlash adjustment off when you're autoguiding. The concept being that PHD figures out the backlash and compensates for it. MaximDL has a similar feature.

Try upping your guide rate to give it a little more 'umph'.
Chris

#19 Madratter

Madratter

    Fly Me to the Moon

  • *****
  • Posts: 6842
  • Joined: 14 Jan 2013

Posted 07 May 2013 - 04:01 AM

I had already upped the guide rate on the declination axis to 99%, which is as high as it will go.

#20 rmollise

rmollise

    Hubble

  • *****
  • Posts: 15704
  • Joined: 06 Jul 2007

Posted 07 May 2013 - 07:29 AM

Well, I learned a whole lot by taking the motor cover off and watching what is happening.

It turns out that if the declination backlash settings are high, at least on my mount, the motor will make an initial big leap, and then do nothing for a while. However, at smaller backlash settings this does not occur. I have moved my backlash settings back down to 5.

I don't know if this will help with my sawtooth behavior when guiding or not.

I suspect not, unless in fact it is switching directions. I don't think that is the case. But either way, this needed fixing.

Watching the motor, I do not think I am seeing cogging, at least as I have seen it described.


OK... If you are imaging, it's advisable to turn all backlash compensation OFF. ;)

I did run mine at .90 guide rate, but you might want to back off from that a bit if possible.

#21 Madratter

Madratter

    Fly Me to the Moon

  • *****
  • Posts: 6842
  • Joined: 14 Jan 2013

Posted 07 May 2013 - 09:58 AM

I originally had my backlash compensation entirely off, but was having problems. I then turned backlash on to see if it would fix it. From what I saw last night, at the levels I had it, I clearly made things worse, not better.

At the time I originally put it in, my scope was badly out of balance. Now that my scope is balanced correctly, it is certainly worth turning it off and seeing what happens. Thanks Rod.

#22 Peter in Reno

Peter in Reno

    Fly Me to the Moon

  • *****
  • Posts: 6079
  • Joined: 15 Jul 2008
  • Loc: Reno, NV

Posted 07 May 2013 - 10:25 AM

Also, and possibly related, when calibrating with PHD, I get the movement along the N axis, but it fails to return very far at all doing the S calibration. Again, I thought that was being caused by weight balance issues, but that does not appear to be the case. I have tried up backlash settings to around 30, but that does not change things.


Increasing Max Dec Duration will help move the guide star in S direction during calibration. If you increase Max Dec Duration to 1500 - 2000msec, you should see the S movement during calibration. It may not move all the way back to original position but close. Turn off Dec backlash compensation in your hand controller. If you do not see S movement at all using large Max Dec Duration, then your mount may have issues.

Think of Max Dec Duration as a backlash compensation. Celestron mounts are notorious for excessive Dec backlash. My CPC0800 had it and increasing Max Dec Duration (1500msec in my case) helped quite a bit that I was able to guide up to 30 minutes per sub.

PHD's Calibration Step Size is depended on focal length of your scope. Don't use it to help reduce backlash during calibration. Increasing calibration step size too much can cause the guide star to jump out of guide camera's FOV and lose the guide star. Adjust calibration step size so that the total steps per direction is about between 12 and 20. The longer focal length of your scope the shorter the calibration step size.

Peter

#23 Madratter

Madratter

    Fly Me to the Moon

  • *****
  • Posts: 6842
  • Joined: 14 Jan 2013

Posted 07 May 2013 - 12:15 PM

Thanks Peter. I'll give that a try and see what happens.






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics