Hi,
Have finally had some clear skies to do some checks. Headline - it was movement of the primary mirror.
Being new to Astrophotography and particularly to Newtonians this issue was difficult to resolve. I found surprisingly little to help online so I have logged my experience here in case it benefits someone else. At the outset having given it some thought and consideration of comments above I was aware that potential sources of the problem included:
Tilt
Camera sensor
Focuser base
Collimation
Secondary mirror - Alignment/height, Rotation and Angle
Primary mirror
Reflections - possibly in the
Telescope tube
Focuser or Comma corrector
Sag - most likely in the
Telescope tube itself (flex)
Focuser tube,
Comma corrector connection to the focuser (eyepiece holder)
Flexure between the main telescope and guide scope, or
Mirror Flop - primary
This process took nearly 3 weeks partly due to cloud cover but also from me going in circles, exploring issues and learning collimation craft. These are the lessons I learned.
Reflections
Some of my imaging was done not that far from a full moon. Also I had previously seen a strange large but faint halo around the bright star in the Horse Head Nebula. However the scope has baffles, some flocking and no unpainted surfaces within the tube so instinct said this was not the problem. As I felt this would be difficult to resolve I continued hoping to find the problem elsewhere.
Tilt
I played with tilt adjustment at both the focuser base and camera. I initially thought these were effectively the same - that alignment errors in either would produce the same outcome and that tilt measured in the image could be resolved with either adjustment. On further consideration I now think they are quite different (happy to be corrected). Tilt at the focuser base is basically a collimation issue. If collimation is good tilt at the focuser base is probably OK and conversely tilt here will appear as bad collimation not "tilt" measured in the image. Camera or lens tilt is different and independent of collimation and tilt in the focuser tube. Assuming perfect collimation, sensor tilt could still be an issue.
That said I ran the scope all one night on a single target. There was no correlation between the 'bad' images and tilt. I got the same tilt readings with the bad images as with the good ones. I would go further and say that low to moderate tilt only marginally effects the images. I concluded that tilt is not the fundamental problem.
Sag and Flexure
I reset and tightened every thing in the image train. I tried to move the centred camera image in OCal by putting pressure on the image train in several directions. I didn't measure it but I would say at least several kilo's of force. I could see minor movement in some planes so inconclusive. My initial experience was that transition from bad images to good images was gradual and associated with approaching the Zenith. This strongly suggested a gravity related problem - likely sag or flexure. At 800mm focal length I did not think this was flexure. Checks with very short exposures which still showed the aberration confirmed this.
Checking another nights images I found that the transitions between good and bad images occurred suddenly between consecutive frames. This was a surprise and had not happened previously where the change was incremental. However it is still consistent with a gravity related issue. In addition I examined images at a Meridian Flip. This significant movement should elicit changes associated with sag, flexure, mirror flop etc. However there was no discernible difference in the images before and after the flip. Tilt readings were also unchanged. This was confusing and seemed to contradict other evidence. There were now several points of evidence against sag and flexure but contradicted by the meridiam flip evidence
Collimation
OCal - I have the second version. I have no previous experience collimating with any other tools. I found the OCal very good for collimating the Secondary mirror. OCal uses the reflected camera image rather than the centre marker on primary mirror for final primary alignment. Mine never clearly focused on this camera reflection making adjustment of the Primary difficult. I got it close but needed to use a de-focused star to get it spot on.
I went around the buoy with collimation several times. In desperation I completely reset the secondary from scratch. This was painful but instructive. It was also probably unnecessary. But in the end I was getting repeatable good collimation of the secondary and Ok collimation of the primary finessed by star collimation. Checks conducted before and after imaging runs indicated that secondary collimation was rock solid. This further supported the conclusion that the problem was not sag. However there did seem to be some movement in primary collimation.
Conclusion
So what was going on? In my mind I had ruled out flexure, tilt, sag, reflections (tentatively) and secondary collimation. That leaves movement of the primary mirror. However as above checks before and after a meridian flip showed no changes. This was a bit confounding but the primary mirror appeared the only option so I pulled it out and took the opportunity to clean it (while testing it rained on me out of a completely clear sky - go figure). I found some very minimal movement was possible in both the lateral plane and in the direction toward the secondary but these did not seem enough to account for my issues. Nonetheless to address lateral movement I placed adhesive rubber strips under the mirror and pieces of paper around the perimeter to remove the very small gaps. I aslo tightened the clips a little to address movement in the other plane - no physical clearance under them this time but still loose enough so as to hopefully avoid mirror pinch. There was a lot of talk online that upgraded springs stop mirror movement. I cannot see this has anything to do with it assuming one is using the locking screws. Nonetheless i did tighten all three adjustment screws thus tightening the springs - which did provide better feel and control while making adjustments.
I put it back together and forgot to redo primary collimation before imaging from 8.00 to midnight when it clouded over. Despite forgetting to collimate, images are good (not perfect but surprisingly close considering). I think I finally have this skinned - thankfully.
Image - single 2min sub, no collimation of primary, average seeing and light overcast.
