Author Topic: The agonies of auto-guidation ...  (Read 556 times)

pregdefciato

  • Active Astronomer
  • ***
  • Posts: 130
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
The agonies of auto-guidation ...
« on: January 08, 2018, 03:23:00 AM »
Greetings from Oregon, USA

A newbie here .. I've never attempted astro-photography but I think I would like to give it a go.

Bear with me please ... I am an electronics engineer and understand closed-loop control systems. I have been reading about PE, mount flexure, mis-alignment errors .. and the requirement of auto-guiding. Why the fuss over the mechanical aspects of a mount, if the errors can be mostly eliminated thru error modeling and a fast closed-loop correction system? What are the limitations of such a system as Celestron's NexGuide autoguider assuming a bright guide star and good signal-noise ratio? Why then agonize over periodic corrections and mechanical issues that can now be essentially eliminated in the closed-loop system? Perhaps I've not thought this out correctly?

Consider allowing the correction/sampling interval to approach zero. In other words, continuous corrections. Are the ST-4 corrections problematic because they are too coarse and discrete? It appears to me if the closed-loop has good response characteristics that all positioning errors could be mostly eliminated. It would certainly relax the mechanical requirements for the mount (poor mount, good corrections, better electronics).

As you can see ... my learning curve is a steep one!

I appreciate any insights offered ...

Dan
Seaside, Oregon



Kenneth Brown

  • Active Astronomer
  • ***
  • Posts: 115
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: The agonies of auto-guidation ...
« Reply #1 on: January 08, 2018, 06:30:18 AM »
Not many agonies here.

Robert Spencer

  • Active Astronomer
  • ***
  • Posts: 127
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: The agonies of auto-guidation ...
« Reply #2 on: January 12, 2018, 05:30:42 AM »
Autoguiders correct tracking errors that have already occurred. For the smallest, most perfect stars it's essential to eliminate those errorsbefore they occur. That's why excellent machining and low PE are still a desirable trait in a mount.

Chris Smale

  • Active Astronomer
  • ***
  • Posts: 129
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: The agonies of auto-guidation ...
« Reply #3 on: January 18, 2018, 02:17:08 AM »
"it's essential to eliminate those errors before they occur."

I was thinking more of 'eliminate those errors as they occur' ... very, very quickly!

ulpehaco

  • Active Astronomer
  • ***
  • Posts: 110
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: The agonies of auto-guidation ...
« Reply #4 on: January 18, 2018, 03:20:02 AM »
The fundamental base for accurate auto-guiding is a good mechanical alignment of your mount, that is, Polar Alignment. Having your mount accurately polar aligned will make your auto-guiding better, no matter what auto-guiding program you use. Accurate polar alignment would require fine mechanical adjustments of your mount by the big mechanical knobs (Altitude and Azimuth), assuming you have a German Equatorial Mount.

Your Go-To alignment relates to pointing accuracy (finding the target). Once found, the polar alignment is necessary for tracking accuracy, allowing long-exposure astrophotography.

Limitations in the gearing mechanism of the mount can be improved by running a periodic error routine, and applying it, to help reduce DEC backlash in the mount. Reducing PE would help the auto-guiding.

If you have a broken leg, and the bone needs surgery to fix the break, do the surgery, so you don't have to use crutches for the rest of your life. In other words, get to the root of the problem, instead of applying a Band-Aid (running a correction program to correct for bad polar alignment). I'm assuming your post does imply mechanical misalignment of the mount would include polar alignment error.

Dan

tioraigenroi

  • Active Astronomer
  • ***
  • Posts: 113
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: The agonies of auto-guidation ...
« Reply #5 on: January 25, 2018, 05:01:29 PM »
Quote
Greetings from Oregon, USA

A newbie here .. I've never attempted astro-photography but I think I would like to give it a go.

Bear with me please ... I am an electronics engineer and understand closed-loop control systems. I have been reading about PE, mount flexure, mis-alignment errors .. and the requirement of auto-guiding. Why the fuss over the mechanical aspects of a mount, if the errors can be mostly eliminated thru error modeling and a fast closed-loop correction system?
The scope + mount + accessories represent a dynamic system mechanically, it is not a monolithic, rigid block. There is a natural frequency limit to the rate at which guide corrections can be applied.

Sean Lee

  • Active Astronomer
  • ***
  • Posts: 124
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: The agonies of auto-guidation ...
« Reply #6 on: January 26, 2018, 06:52:31 AM »
Quote
"it's essential to eliminate those errors before they occur."

I was thinking more of 'eliminate those errors as they occur' ... very, very quickly!


Think of it however you like, but the errors have already occurred and will have some impact on the image.

tradneedcoegen

  • Active Astronomer
  • ***
  • Posts: 119
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: The agonies of auto-guidation ...
« Reply #7 on: January 30, 2018, 06:48:05 AM »
The *main* reason why your approach won't work has been mentioned before, but bears focusing on: Seeing.

When the guide star "moves" because of atmosphere movements, you don't want to move the mount to compensate--because you guarantee the need to move it again next instant. You only want to move the mount when *it* (the mount) has drifted from the correct path of the star.

This is why usually it is recommended that corrections are made less rather than more often, so you don't "chase the seeing." With high quality (premium) mounts, the period between autoguide corrections usually gets longer, to 10 seconds or more.

rissubssimpsat

  • Active Astronomer
  • ***
  • Posts: 107
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: The agonies of auto-guidation ...
« Reply #8 on: January 31, 2018, 03:48:33 AM »
Quote
The *main* reason why your approach won't work has been mentioned before, but bears focusing on: Seeing.

When the guide star "moves" because of atmosphere movements, you don't want to move the mount to compensate--because you guarantee the need to move it again next instant. You only want to move the mount when *it* (the mount) has drifted from the correct path of the star.

This is why usually it is recommended that corrections are made less rather than more often, so you don't "chase the seeing." With high quality (premium) mounts, the period between autoguide corrections usually gets longer, to 10 seconds or more.

In *theory* a guiding system that included both mechanical guide corrections and active optics might be able to reduce the effects of seeing errors. In reality, few amateurs are willing to create a system with nested feedback control loops that could make it happen. It's simpler and cheaper to buy a quality mount with low and predictable PE and then attach the imaging scope as rigidly as possible. PEC and/or encoders coupled with centroid guiding usually gets us to a place where the limitations are the accuracy of polar alignment and the local seeing conditions.

Robert Farley

  • Active Astronomer
  • ***
  • Posts: 126
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: The agonies of auto-guidation ...
« Reply #9 on: January 31, 2018, 05:44:52 AM »
True, I am assuming a fairly high degree of alignment and not the extreme case of no alignment at all. Consider flexure of the system that may change with loading schemes. Difficult to model and more easily controlled via closed-loop.

I appreciate the comments ...

Dan

Jermaine Conner

  • Active Astronomer
  • ***
  • Posts: 111
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: The agonies of auto-guidation ...
« Reply #10 on: February 03, 2018, 04:48:29 AM »
"Consider allowing the correction/sampling interval to approach zero. In other words, continuous corrections. Are the ST-4 corrections problematic because they are too coarse and discrete?"

As has been stated previously, this is not feasible because of seeing. Absolute encoders get over this hurdle because they do not have seeing to contend with. However, for absolute encoders to work, the mount itself must also be precise.

The reality is that you are always working with uncertainty. Uncertainty in where the guide star is (due to seeing) and uncertainty in where the mount is pointing (due to machining). You can only do so much with the uncertainty with the guide star, but you can make the mount more precise, for a price. To fix the uncertainty with the guide star, you can use absolute encoders instead, assuming that your mount's uncertainty is lower or equal to that of the encoders.

It isn't impossible by any means to create a mount that can track natively with less than 1" of error. It just takes an enormous amount of time, skill and testing.

Jason Hillyer

  • Active Astronomer
  • ***
  • Posts: 112
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: The agonies of auto-guidation ...
« Reply #11 on: February 03, 2018, 07:41:24 AM »
Take a look at adaptive optics. For example the SX AO system. LINK

It will do some of what you are shooting for however as has been previously stated the 'system' you are working with is not monolithic and external conditions can be quite variable which means a great deal of complexity that needs to be controlled for. The old engineering maxim comes into play: You can have it fast, well made, or cheap. Pick two. Substitute fast for precise.

Paul Hunt

  • Active Astronomer
  • ***
  • Posts: 124
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: The agonies of auto-guidation ...
« Reply #12 on: February 09, 2018, 10:54:32 AM »
Quote
"it's essential to eliminate those errors before they occur."

I was thinking more of 'eliminate those errors as they occur' ... very, very quickly!

That is the basis of Periodic Error Correction and the use of absolute encoders. The other factor is that tube and mount flexure and mirror movement plus atmospheric conditions and wind acting upon the mount will induce non periodic errors, aside from those in the drive.

I find that with most images it is necessary to use autoguider exposures on the order of one second or more; add to that the time to download, process and send a guide correction bask to mount.

Edit: I see that I posted before reading that some of these concerns have been addressed.