OTA Hardware or Star Alignment Challenge?

rwatters

Well-known member
To fess up right at the start of this thread, I have 90% convinced myself that I have a focuser problem with my recently acquired, second hand, Tak FSQ106 EDX III. However, with a view to eliminating doubt and to look at the issue from another angle, I would like to treat this as a star alignment challenge and see what can be gleaned from the process.

I have captured a number of test files which are saved in a drop box folder here. Image data was captured with SGP Pro and I made sure to check that the focus was as close as possible. Guiding was running fine at .5" for the duration of the testing. Image train is Tak FSQ106 -> M72-M68 adapter -> Gerd Numen M68 tilt adapter -> M68 extension -> ZWO M68 OAG -> ZWO 2" x 7 EFW -> ZWO ASI6200. The ASI 6200 is a full frame CMOS sensor with 3.76 micron pixels.

The first thing that jumped out at me indicating that there was an issue was to do with registering images that were taken pre and post meridian flip. Upon further inspection I can see what appears to be an elongation in the stars in the centre of the image. This is not consistent across the field which can be seen from the images taken with the Aberration Inspector script in PI (see images in drop box folder). Indications are that there is either focuser sag or something else which is why I'm looking at replacing the focuser. Though, as stated above, I have a feeling that by approaching this as an image registration challenge I may learn something useful.

As part of the testing I took one image with the OTA west pointing east and then a second image with the OTA east looking west. The object was near the meridian at the time. I then used the star alignment tool to register the two images and by blinking them we can see that there is a problem with alignment. It is almost like the star alignment process has aligned properly on the centroid of the star which causes the elongations to not match up properly in the registered images. If this is confirmed to be the case then I will stop pursuing this line of enquiry and concentrate on the OTA performance.

The two images concerned in this test are:
* Test_180s_Lum_Gn_0_OTA_west_20210122_01_Annotated
* Test_180s_Lum_Gn_0_OTA_east_20210122_02_registered_Annotated
The first image was taken with the OTA west looking east. The second post meridian flip with the OTA east looking west. I then used that star alignment tool to register the two images (image *_01 as the reference) and annotated them using the render script. This allowed me to compare the 'aberrations' that I'm seeing against the RA axis. To my eye it confirmed that the issue is not to do with guiding but I'm happy to challenged on this (or any of my theories for that matter :) ).

Here is a link to a video file in the drop box folder which is a ten second snapshot of the blink tool.

So my question to this forum is, is there anything to be gained by looking at this as an image registration challenge?

Thanks and clear skies,
Rodney
 
It doesn't look as though there is any problem with registration - all the small stars are in rock-solid alignment. The blink displacements are for the brighter stars. For these stars, there seems to be a systematic displacement between the images, which is more or less the same across the whole frame. Since I assume the optical centre of these stars must be in common alignment with the smaller stars, the star images must be asymmetric. I'm not sure how focuser alignment affects images, but that certainly looks like a possibility.
 
i agree, the smaller stars are registered properly. i wonder what happens if you try with a synthetic reference (made by StarGenerator) and then use thin plate splines in staralignment. on the other hand if SA is correctly finding the centroids (which is extremely likely) then this won't help.

this is a pretty interesting aberration - i don't think i've ever seen anything like it. i wonder if it has to do with tilt somewhere in the imaging train. these FSQs are awesome when they are working right but it seems like there's a lot to go wrong with collimation due to the petzval design. and then in the USA the only place to send it for repairs is that texas nautical place which seems to be hit and miss.

rob
 
It doesn't look as though there is any problem with registration - all the small stars are in rock-solid alignment. The blink displacements are for the brighter stars. For these stars, there seems to be a systematic displacement between the images, which is more or less the same across the whole frame. Since I assume the optical centre of these stars must be in common alignment with the smaller stars, the star images must be asymmetric. I'm not sure how focuser alignment affects images, but that certainly looks like a possibility.

Thanks for your response. You seem to be coming to the same conclusion as me which basically is that the data is at fault here.
 
Last edited:
i agree, the smaller stars are registered properly. i wonder what happens if you try with a synthetic reference (made by StarGenerator) and then use thin plate splines in staralignment. on the other hand if SA is correctly finding the centroids (which is extremely likely) then this won't help.

this is a pretty interesting aberration - i don't think i've ever seen anything like it. i wonder if it has to do with tilt somewhere in the imaging train. these FSQs are awesome when they are working right but it seems like there's a lot to go wrong with collimation due to the petzval design. and then in the USA the only place to send it for repairs is that texas nautical place which seems to be hit and miss.

rob
Thanks Rob, I’ve decided to take a chance and replace the focuser on this OTA with a Feather Touch unit. Other testing that I have done indicates tilt and/or some sort of movement on the focuser. You are right though in that this sort of aberration is rather unusual and it doesn’t jump out at you until attempting to register the images.
 
there's a lot to go wrong with collimation due to the petzval design
I'd missed that this was a petzval scope. I guess a slight collimation error (invisible on the smaller stars), rotated by 180deg with the flip might just explain this.
 
The idea of trying to collimate the petzval cell itself would scare me to death, but I guess collimation of the focuser (and CAA if you have one) are the first targets, and may be easier to achieve (does the FSQ have any focuser collimation adjustment?).
 
Back
Top