12 replies to this topic

[orlyandico]

Hello all,

 

I've noticed these annoying egg stars in my individual subs. They are not present in 4-8 second images so they are certainly guiding artifacts.

 

Here's a screenshot. The guiding is generally good, and while I'm not perfectly polar-aligned, it is very very close (I use a Polemaster).

 

Any ideas what's causing this egginess?

 

 

Note that the weird shaped guide star is because of the OAG, and this is from an Epsilon where the off-axis stars are weird and wonky. But I get the same egg stars with my AP130.

Edited by orlyandico, 14 September 2024 - 08:00 PM.

[idclimber]

Can you upload one of the subs to a shared drive and post a link? I would also include one of the short subs if you can. 

[Linwood]

I would start by being sure it is tracking.  Look carefully at a raw sub and the whole frame.  Tracking will look the same throughout the image, optical effects may vary across the frame (tilt is an exception but its strength will generally vary across the frame also). 

 

If you are sure it is tracking it may be useful to see if it is consistent, always in the same orientation throughout the session.  Plate solve an image and figure out if it's along the RA axis or the DEC axis alone.  This can give indication where to look as well. 

 

Diagnosing and fixing would likely require a lot more info, a copy of the guide log, a long-ish run of Guide assistant, etc.  But do some digging first to ensure you are on the right track, and narrow down the scope of the problem.

[orlyandico]

Here's a 4-second sub (50MB): https://drive.google...?usp=drive_link

 

and a 120-second sub taken a minute before: https://drive.google...?usp=drive_link

 

I reviewed my subs from 2 nights ago, same setup, and the stars were round in the guided subs. So it's not consistent.

 

Also: the stars off-centre are going to look funny, as this is a 27-year old pre-ED Epsilon that's not 100% collimated.

[Linwood]

To me the shapes look pretty much the same.   Which would point to optics not tracking.  Though with optics that bad as you get away from the center it makes it a bit harder. 

 

[orlyandico]

Here's a 2-minute sub from the AP130, all else being equal. PI reports an eccentricity of 0.52 which is not fantastic.

 

https://drive.google...?usp=drive_link

 

The 4-second Epsilon sub was eccentricity of 0.63 which is pretty bad, although I've had some with eccentricity as low as 0.53.

 

Could it possibly be the ZWO tilt plate? I never made any effort to adjust for tilt, I just backed off the grub screws so the tilt plate is fully seated.

Edited by orlyandico, 14 September 2024 - 09:20 PM.

[Linwood]

There are a number of ways to analyze for tilt, ASTAP gives a good indication.  

 

If all else fails, with everything else the same, take some subs, rotate 90 degrees, take some subs.  Tilt will stay the same, tracking should rotate 90 degrees. 

 

Finding the causes requires digging a bit.

[acrh2]

It looks like your back focus is incorrect. You should use 56.2 mm from the corrector lens per the OTA manual. Plus more if you are using filters.

[orlyandico]

I checked with ASTAP, there was negligible tilt with the AP130 but 12% with the Epsilon.. which is probably due to the low-quality focuser that Takahashi puts on these things.. there's obvious sag in the drawtube unless the locking screw is really screwed down tight.

 

I had oriented the OTA so that the focuser was parallel to the mount counterweight shaft, but M27's location meant the camera was hanging about 30 degrees from the horizontal.. I don't want to keep rotating the OTA in the rings for every different DSO, because that will rotate the diffraction spikes relative to the stars.

Edited by orlyandico, 14 September 2024 - 09:32 PM.

[idclimber]

I concur with acrh2 that the 120" sub indicates you need more backspace. The indicator is the radial pattern in the four corners. You may also have some collimation issue, but you need to work on the backspace first. 

[orlyandico]

Good information, thanks. I had to adjust the backspacing on the Orion 3" flattener I'm using on my AP130 many times. I didn't notice it in the case of the Epsilon because I thought the corners were because of coma, and I didn't think the same rules (radial pattern too close, circular pattern too far) also applied to reflectors.

[Linwood]

Good information, thanks. I had to adjust the backspacing on the Orion 3" flattener I'm using on my AP130 many times. I didn't notice it in the case of the Epsilon because I thought the corners were because of coma, and I didn't think the same rules (radial pattern too close, circular pattern too far) also applied to reflectors.

It mostly applies when the reflector has a flattener or reducer, e.g. the Celestron Edge's have one built in.  But the Epsilon is a bit of an odd duck so I won't speculate.  However, I'm sticking with my original -- tracking induced egg shapes will be uniform across the frame.  Radial are almost certainly optics (like backfocus).  Same direction but different strengths, notably weaker in middle or edge than opposite, maybe tilt.   The 12% mentioned is pretty mild for tilt per ASTAP's measurements.   When you have a mixture of these causes I recommend a high proof bourbon before analysis. 

[orlyandico]

I just had a very interesting discovery... the images above were with a backspacing of approximately 57.5mm which is already greater than Tak's recommended 56.2mm

 

Based on the advice on this thread, I increased the backspacing to 59.5mm - and the corner egg stars became much worse. In fact coma showed itself in its classic form (comet-shaped stars with the tail pointing away from the centre).

 

It looks like the curvature of this old Epsilon and corrector is actually the reverse of what we normally expect!

 

The obvious solution is to go for a backspacing of exactly 56.2mm or as close to it as I can get..

 

I also changed the focuser to a Moonlite 2.5 CRL and collimated the entire train to within an inch of its life. I can now get 3.62 pixel HFR (measured by NINA) which is quite close to what I could get with my AP130 in my (somewhat hazy) skies. So it looks like even this circa 1997 pre-digital Epsilon can produce acceptable results..