Wavelength dependant reflection ghosts

Thanks Tim.

Now you have explained it I can see that the "glass" part is the outer edge of the support and the inside edge is milled away over the CCD. Doug was right, you can't diagnose very much from an image, easy to be led astray.

William.

 

too bad, this would have neatly explained a problem that has stumped everyone so far. Only a couple of more things to re-check before I will have exhausted or eliminated everything but the camera/fw. to re-cap here is what I have looked at in no particular order:

1. Collimation verified and adjusted multiple times, on the plus side I really have the process down and the stars are perfect little round points across the FOV, analysis has shown the secondary mirror distance is withing 0.5 mm of perfect (1600 mm focal length)
2. All even remotely reflective surfaces have been treated with flat black chalkboard paint - my telescope and all the parts in the imaging chain look like the inside of a black hole
3. Flattener removed with no effect on the ghosts, also Chris at OPT kindly swapped flatteners out with me to make sure it was not a bad flattener - I am going to redo this test one more time at Doug's request
4. Tested my telescope with the flattener and my Canon T5i, no ghosting seen, very nice images BUT no filters in the imaging chain so its not exactly apples and apples
5. Checked the orientation of my Baader filters - no noticeable difference observed in the ghosts - currently the filters are in the recommended orientation with the filter "shiny side" facing the stars
6. Imaged with the filter wheel set to the open slot - in this configuration there is no ghosting visible
7. removed the OAG assembly and tested it - ghosting still present in all filters (RGBL, Ha, O-III, and S-II)
8. Cleaned my primary and secondary mirrors
9. Mounted a filter at a tilt in the filter wheel to make sure the surface is not parallel to the CCD imaging plane - the ghosts were slightly distorted (no longer perfectly round) but still there and I guess it confirms that the filters are involved
10. checked for frosting (none visible) and operated with the cooler off - ghosting still happens
11. tested with an "external" filter an Astronomik UHC, with FW switched to open slot the ghosting was seen but the aperture ghosts were larger due to the filter being further away from the focal plane, which makes sense

I am going to do one more test after I get a 2" visual back for my Celestron C8 to mount the STF-8300/FW/OAG on it and try with this completely different scope - it it still does it then I will have eliminated the 8" RC and its extensions/ feathertouch focuser

the 8" RC with the STF-8300 is taking beautiful pictures as long as there are no bright stars in the FOV (Vega, Capella, Alnitak, and Sirius have all shown the problem) but nobody else seems to see this with this combination of camera/scope

Unless I am missing something (please let me know !) by process of elimination if the camera still does it in the C8 then there are two possibilities - either I have 8 filters from two different manufacturers with faulty AR coatings, or there is something wrong with the window/CCD in the STF-8300M

 

The only other possibility I could think of is that the KAF 8300 is available from OnSemi in several versions with a plain glass cover slip, not AR coated, and it is just possible that one slipped through the supply chain somewhere. Problem is the part number of the CCD is only visible from the rear of the CCD chip and the camera would have to go back to SBIG for them to visually verify that the CCD part # / serial # matches their production records.

Looking at the QE figures for plain glass slip V AR glass slip the difference is only 2% which doesn't seem enough to account for the strength of the reflections that appear in your images.

Good luck with the test on a different telescope, hope the problem is soon resolved for you.

 

The single images are very stretched to show the reflections - however since they are correlated they add up when stacked - I discovered this issue imaging the flame nebula and for 5 hours of data the ghosts produced artifacts that were almost as bright as the nebula itself

 

Has been terrible conditions for looking at stars but I setup my artificial star (LED flashlight with an eyepiece cap in front of it) Here is the STF-8300M and Filter wheel combo looking through my Celestron C8. NO flattener installed and the OAG has been removed.
Ghosts seen on all fiters (Ha image attached) and Not seen when shooting through the open filter slot (no filter image attached)

 

the RC scope is well collimated, as is the RC, to be honest the collimation appears to have zero effect on the issue I am seeing - I focused the C8 using a Bahtinov mask and took a picture to document, obviously my artificial star isn't quite a pinpoint, but again that doesn't seem to really matter either - the reflections seem to be dependent on only the intensity of the source. There are no shared optical components shared between the 8 SC and the 8 RC other than the camera/fw. I was pretty sure the telescope itself was exonerated when I shot some clean pictures with my Canon T5i (with the flattener too) but this was the final test to make sure it wasn't some strange optical issue with the mirrors in the RC.

 

I was able to see it on two different telescopes, an f/8 Ritchey-Chretien and an f/10 Schmidt Cassegrain (a C8) and with all of the Baader filters RGBL, Ha, OIII, and SII. It shows up on long integrations on bright stars - you can see it in a 2 min integration if you stretch the image in PixInsight, because the artifacts are correlated from image to image over a long session with many 2 minute subs integrated the effect will be come quite pronounced. if you look at shorter integration (like 10 or 20 seconds) it might be very subtle. I used a white LED flashlight as a source behind a pinhole in an eyepiece cap.

maybe this info will help you duplicated the problem

 

A typical cool white florescent bulb has very little emission in the band where that red filter is, a white LED will have a much broader spectrum, you might have to integrate much longer using the CF tube

 

thank you Doug, I can't wait to see the results !

 

Any Update? I am also getting artifacts when bright stars are in the FOV so am curious what the solution is or good techniques to troubleshoot.
Eric

 

Eric, post up an image of what you are seeing