Hi Paul

Well I am having some trouble trying to collimate my RC with your single star collimation tool. I have a 14.5 RCOS using a SBIG STL11000 CCD camera. I found a bright star and defocused it to look like you example in the instructions. However, every time I try to adjust the mirror collimation I tend to get CCDI telling me to continue to move the mirror in the wrong direction. I finally had to just use the image to get the collimation back to near center. I have about 200 pixels being used for the defocus star with about 30,000-45,000 levels on pixels. I hope to try it again tonight, but not having a lot of success here...

I then tried the multi star tool and got errors like "stars not evenly distributed" many times - after about 30 min of hunting around star fields I finally got it accept a star field (occasionally) to provide me some numbers. Is there a setting I should be adjusting?

Thanks for any help

Jason

Hi Jason,

Hard to tell about the single star collimation without seeing the actual raw images.

A few things to make sure when doing a single star collimation:

1. Try to stretch the image histogram hard, and see if the star is a double. A reasonably bright companion will consistently skew the results in one direction.

2. Check that the star is not saturating the sensor (try shorter exposures and lower ADU levels). An ABG chip can saturate the pixels and not have an obvious artifact to show this. Stretch the histogram hard and see if the outside star shape appears circular, and if there are any long streaks or other distortions emanating from it -- these will also affect the collimation calculation.

3. Make sure the star is well centered on the chip for each measurement. If you sub-frame the chip, make sure the subframe itself is also well-centered .

4. Try sub-framing the image to cut out possible contamination from near-by stars. These may be hard to see in the image due to the brigh star in the center, but they'll still affect the calculation. As in #3, center the sub-frame on the chip.

5. Auto-dark subtract the images to reduce hot pixels

6. Try a dimmer star and bin the chip 2x2 or even 3x3 (keep the ADU levels below about 45,000 adu. This will help with reducing the noise.


For the multi-star collimation method, good focus and a rich star field are needed. If you get warnings about too few stars or stars not spread out enough, try increasing exposure, increase binning, or use image averaging under Real-Time->Images to Average setting.

Regards,

-Paul

Hi Paul

ALl great suggestions - I hope to work on this tonight - clear skies predicted

Jason

Paul

had mix results over the weekend.. I will send representative images for both multi-star collimation and for single star collimation... I find i can do better with my eye than the tool - perhaps it is just my use of the CCDI

Jason

Hi Jason,

I'll see if I can spot what's going on from the images.

Regards,

-Paul

paul

atached are two images:

single and multi star fit files

Thanks
Jason

single-star_collimation_1.fit (79 KB, 29 downloads)

Paul

multi star collimation file fit is too large (5 MB) - tried to email - can you provide me email address

thanks
Jason

Jason,

Please email to:

paul [at] pk.darkhorizons.org

Thanks!

-Paul

Jason,

It appears that this star has a close companion. Here's the image stretched to show it:



And, here's the collimation viewer display produced from this image that obviously shows a reading skewed in the same direction as the second star:



Collimation viewer is very sensitive, and any extra flux that doesn't belong to the star being measured will interfere with the reading. It appears that that's what has happened here.

Regards,

-Paul

Hi Paul

this is exactly what I needed - will retry the single star and look again for any other stars near by - I have also emailed you multi star fit files...

thanks
Jason

Hi Jason,

I received your multi-star collimation image. Everything appears to be OK, except that there are many areas in the image with no stars. What that means is that CCDI has to interpolate (or make an intelligent guess) as to what star shapes and sizes should be in those areas. Generally, this is not a fatal problem unless a really large portion of an image has no stars in it.

Collimation result shows that collimation is very nearly perfect, with a very slight tilt to the imaging plane towards the left.



You have a few choices:

1. Ignore the CCDI warning (click on Ignore checkbox on the message) and see if the collimation result is repeatable.

2. Find a more densely populated star field / open cluster

3. Try to average a few images together (your exposure is already on the long side at 30 seconds, so averaging multiple images will probably take too long).

So, I'd recommed you try to find a better popualted star cluster, or simply use the single star collimation tool.

Just make sure there are no other stars near-by. To do this, you can focus the telescope, take a long exposure (say 10 seconds) and see if any stars show up near the main star.


Regards,

-Paul

This message has been edited. Last edited by: Paul Kanevsky,

msc.jpg (35 KB, 133 downloads)