Hollow stars and coma problem

I suspect the problem is reflections from the filter. I don't think is has anything to to with collimation, spacing or flexure.

It's a really bad case.

Edit: Sorry. I was looking at the wrong thing. I guess it is spacing plus some collimation error.

Thanks everybody for the kind answers...I cannot believe I could be so carelss...the good thing about discussions is that they trigger ideas and thoughts perhaps buried in one's mind...In wanted to double check that I have the spacers assembled in the right way and I just realised that I have the filter wheel assembled the wrong way round...the filters don't sit precisely in the middle of the EFW so I suppose that this has an effect on possible reflections; moreover I remember spending some time trying to find the "right" side of each of the filters, so on top of everything I've used the filters in wrong orientations. I'm not sure this error explains all the problems I have... perhaps I still need working on the correct spacing 

Hi Michele,

From your description, you're explicitly stating that you are using a coma corrector with the ZWO. For the Canon you don't explicitly say it but I assume you're also using it with the Canon, right?

The back focus of the ZWO you have is 17.5mm measured from the beginning of the T2-adapter (I assume you use the T2). So to get the 55 of the Canon, you need some more distance but your filter wheel might take quite some space already, so you shouldn't need a lot.

The distortion of the stars is radial. The stars in the center look good. It's not coma from a bad collimation. If it were, you'd also see it in the center of the image.

CS!

Björn

Erlend Langsrud:
I think the collimation is quite good. The aberrations are quite symmetric around the center.

No wonder you get large reflections if you mountet the filters the wrong way. Narrowband filters are often mirror-like on one side.

I agree that the abberrations look like off axis astigmatism and not coma, but in either case, the spacing must be way off.

The optimum orientation of thin film filters seems to be a point of widespread confusion so allow me to explain this in some detail.  The key issue centers on how the filter might generate stray light, which normally comes from light reflected from the sensor itself.  Strays can also come from the other components between the filter and the sensor such as a cover glass, but here I'm going to focus just on the stray light coming from a bounce off of the sensor surface.  Since the sensor surface sits precisely at the image plane, the size of any stray will be determined by the separation between the filter and the sensor along with the focal ratio of the telescope.  The larger the separation, the larger the stray.  As the stray gets bigger, the irradiance of the stray decreases making it less of a problem.

We can look at how much optical power gets dumped into this stray reflection pretty easily.  Each surface has three things that it can do to an incident beam.  It can reflect it, transmit it, or absorb it.  No surface is perfect so absorption is never identically zero but for most common optical coatings the absorption values are very small, so for this calculation we are going to ignore absorption and consider only reflection and transmission. Under these circumstances, the conservation of energy requires that T=1-R.  This reduces the task of computing the transmitted power to a matter of simple book keeping.

The figure below shows two configurations--one for the thin film (shiny side) up and the other for the thin film down (closest to the sensor).  Here, we simply consider the amount of optical power that makes it into the stray from a bounce from the thin film layer.  I've picked fairly typical values for the performance of the thin film coating and the AR coating on the other side. I'm also assuming a reflectivity of the sensor to be roughly 40% for a typical sensor but that will vary with the type of sensor, fill factor, and coatings on a micro lens array.  Either way, that value won't change the relative results for our comparison.  As you can see, in this case, the orientation of the filter does indeed make a small difference in the total power going into the stray.  The ratio of the differences is driven by a factor of T2^2 (98%).   So, putting the shiny side down puts a little more power into the stray.  Furthermore,  the size of the stray will be a bit smaller in size, which will increase the irradiance.  How much so depends on the focal ratio of the telescope.  In this case, placing the shiny side up is preferable.



But wait a minute, we aren't done yet!  This first calculation only shows the effect of the single stray from the thin film coating.  What happens if we add the stray light reflected from the AR coated side of the filter?  This second diagram (below) shows that situation.  In this case, we can see that the power going into the two (mostly) overlapping strays is nearly identical at 2.2% of the total incident power.  So which orientation is better?  The answer is that they are both very close; however, there is still a very small benefit to putting the shiny side up simply because it will make the stray reflection slightly bigger.  Since the optical power is nearly the same, spreading that power over a slightly larger area will reduce the irradiance by a small amount and that's what counts.  However most filters are pretty thin (commonly 3 mm) so in practical terms, you'll be very hard pressed to see much of a difference between the two orientations.

Conclusion:  Filter orientation won't make much difference.