SC3 binning and guiding

Discussion in 'Guiding and Adaptive Optics - StarChaser and AO' started by mike shade, Feb 26, 2021.

  1. mike shade

    mike shade Cyanogen Customer

    Joined:
    Sep 14, 2015
    Messages:
    75
    Is there a way to bin my SC3 more than 2x2? At my focal length, 2940mm binned 2x2 gives me about .65"/pixel or so. Chasing the seeing is putting it mildly. This is very high resolution, and after three nights trying to get this thing to guide-it does not, at least satisfactorily. It would be nice to be able to bin this 3x3 or more, if possible. I have worked with the settings in Maxim, delay after correction, aggressiveness, minimum move and the other usual suspects. Both ASCOM direct and guider relay have bother been tried. There is slight vignetting in the field from the pick off mirror, it may be able to be moved a little bit (in the field maybe 10%+/-). Focus seems OK for an off axis guider. My hope was to be able to run ACP by now, there is no way this camera and guider are ready for that. There are several issues:

    *It like to calibrate on a hot pixel or something, even when there is a bright star in the field.
    *It likes to "half calibrate", where it will calibrate in one portion of the "L", but the other goes off at an odd angle, loses the star or something, or does a V calibration, or some other unusual design.
    *It oscillates in RA...runs up...then runs down...then back up...and so on. Images are useless. Unguided images of 60", 90", 120" show pretty round stars. Yes I do need to guide the system.

    Less than a week ago, the imaging system was working just fine, guiding was good-(CDK17/AP1600GTO/STL6303E). This is my reference for performance of what the telescope and mount are capable of.
     
  2. Doug

    Doug Staff Member

    Joined:
    Sep 25, 2014
    Messages:
    8,171
    Unfortunately no, 2x2 is the sensor's binning limit. CMOS sensors in general have rather limited binning capabilities compared to CCD.

    It is possible, of course, to introduce software binning in the driver. If you think that is an important option we could add that to our development queue. I can't make any guarantees on when that might become available, though, as we have many priorities!
     
  3. mike shade

    mike shade Cyanogen Customer

    Joined:
    Sep 14, 2015
    Messages:
    75
    This is a problem I think. I suspect that given the resolution, even binned 2x2 (about .65"/pixel), it is going to be virtually impossible to get this system to guide accurately. A 17" telescope with a 2940mm fl is going to be impacted be seeing as you are aware, especially the seeing I have here at my site in AZ (average about 2" as measured by a SBIG seeing monitor). It is going to chase the seeing and while I might be able to mitigate it somewhat, this is very high resolution for guiding. I binned the STL's guide sensor 3x3 and it helped although the ST237 has larger native pixels. I was able to get it to guide reasonably well regularly. I also assumed when I purchased the system (Aluma AC4040, FW7, filters, and SC3) that the SC3 would be able to be binned adequately even with the 4.8 micron pixels. Had I known that this was not the case, I would not have purchased it. In my review of the literature, I did not see this mentioned. The SC3 is the only guiding option.

    Thus far I have spent three nights working on the guiding. It oscillates in RA (somewhat in DEC as well), the stars are more on the line spectrum, not the very slightly ovoid stars see in unguided images. As I mentioned it can certainly do an unguided image with stars much tighter than when guided. Nothing I have tried has made an appreciable difference; increasing the minimum move, delay after correction, aggressiveness. This is not to mention the other issues, raised above.

    I have a few things to double check mechanically: cable drag, balance, gear mesh, possible a new PEC curve. Again, a week ago this system worked satisfactorily with the STL camera. PEC curves for example don't "go south" in a week.

    However my suspicion is that it will never guide satisfactorily as the guider has such high resolution. In being at this site for 22 years and using a number of self-guiding cameras, I know from experience this is unlikely to be workable. I have done this before and high resolution guiding has not been workable as I have seen this behavior before with different cameras, telescopes, and mounts. I have had to lower the resolution on every guiding system I have used with AP1200s, MEs, and the 1600.

    I will have to spend yet another night on this tonight, so far approaching 20 hours screwing around with it.
     
  4. Doug

    Doug Staff Member

    Joined:
    Sep 25, 2014
    Messages:
    8,171
    There are lots of things you can do to impact guiding. Of course the first thing is to simply dial back aggressiveness, which effectively creates a rolling average effect. There are also minimum/maximum move settings in the Advanced dialog. Anti-stiction can also be useful in some situations.
     
  5. mike shade

    mike shade Cyanogen Customer

    Joined:
    Sep 14, 2015
    Messages:
    75
    Yes I do quite understand that there numerous variables. I have dialed back aggressiveness, increased the minimum move, increased the delay after correction. Also using a short guider exposure. None of this has made much of a difference. As I offered, I have had similar experiences with the two previous cameras ST10XE and the STL6303E. In both cases, had to bin as much as possible for guiding and this made a significant difference as it decreased the sensitivity to seeing variations with accurate guiding being the result. Unfortunately, I can't do this with the SC3.

    I will work on this tonight, as I offered, and there are a few things I can double check.

    Unless someone has a clever solution, and if I can't get a handle on it, then the system will go up for sale.
     
  6. Doug

    Doug Staff Member

    Joined:
    Sep 25, 2014
    Messages:
    8,171
    We've talked it over and we're going to implement software binning in the driver. It will be able to do 3x3 and 4x4 binning. The 4x4 will be 2x2 in hardware and 2x2 in software.

    I've asked the team to prioritize it. I assume you'd be interested in beta testing the updated driver?
     
  7. mike shade

    mike shade Cyanogen Customer

    Joined:
    Sep 14, 2015
    Messages:
    75
    Yes, I would beta test it. One variable. My wife and I close on a house in rural New Mexico March 25. Oddly enough, will be setting up a larger observatory there with room for 5 scopes, at least three of them remote. It is a good spot at 7450' elevation, on 25 acres. So I will be in SE Arizona for another month, and sporadically after that. I am not expecting the new observatory to be built for several months.

    In sum, if the driver is going to be soon I can test it, if it is going to be in several months, will not be able to do so.

    As I offered, there are a few mechanical things I am going to verify on the system including balance, cable drag, gear mesh, possible a PEC curve. While I think these are set, will double check. Can play with software parameters, but as I have been doing so for some time...

    Two final questions: How does one tell if there is a cover on the pickup mirror? I assume it is blatantly obvious. Second, could the guider be mounted 180 degrees off? Meaning it thinks that X- is actually X+ or some weird thing? I wouldn't think so as it will when it has a mind to calibrate, but at this point.

    Thank you and to the team.
     
  8. Doug

    Doug Staff Member

    Joined:
    Sep 25, 2014
    Messages:
    8,171
    We're going to expedite the feature. The driver already contains code for doing software binning in certain circumstances; we just need to add "additional circumstances".

    There is no cover. Make sure there isn't a plastic protective sheet on the mirror (sometimes the supplier includes them, sometimes not - we normally remove them). The only other way the chip might be covered up is if it's being vignetted; in this case you may need to move the mirror inwards.

    Yes the guider can potentially be mounted in four different orientations, but our software doesn't care about +/- directions - that is handled by the guider calibration process.

    When I'm not completely sure what I'm seeing, I like to point the telescope at the moon. It's easy to find, and it helps you figure out things like the orientation of the guide chip relative to the main camera.
     
  9. mike shade

    mike shade Cyanogen Customer

    Joined:
    Sep 14, 2015
    Messages:
    75
    OK, so when I shoot an image through the main camera, I see the dark outline of the mirror on the lower part of the image, maybe 10% to 15% into the frame. It is definitely there. When you say move inward, inward being towards the opening or towards the outer edge of the SC3? I'm not sure at what point the mirror is properly positioned in the light path. I assumed that minimizing this is the goal. Having no vignetting would be the ideal, but my guess is that will limit any availability of guide stars so some is necessary. Question is how much.

    Thaks,
     
  10. Colin Haig

    Colin Haig Staff Member

    Joined:
    Oct 27, 2014
    Messages:
    4,925
    Location:
    Earth
    With the huge image circle the CDK17 has, there should be plenty of room for starlight to hit the guide sensor without vignetting the main chip.

    Doug's "inward" is like a turtle retracting head into its shell. e.g. away from the centre of the light cone, and in toward the metal work.

    You should be able to slide the pickoff mirror assembly back into its housing more - unblocking the light path to the main sensor.

    Then turn the focus knob counter-clockwise on the SC-3 to move the CMOS guide sensor in toward the pickoff mirror.

    CW = Move sensor away from the pickoff mirror.
    CCW = Move sensor toward the pickoff mirror.

    The right point is when the main sensor is not vignetted and the guide sensor is seeing stars. They will be imperfect (not nice round ones) due to the pick-off mirror, but should be fairly tight.
     
  11. mike shade

    mike shade Cyanogen Customer

    Joined:
    Sep 14, 2015
    Messages:
    75
    Thank you Collin, among other things will move the sensor a bit and see how that goes.
     

Share This Page