Using Flats as a Tilt guide

[AlistairW]

I was wondering if there was any merit in this approach I was trying this morning.

Let us assume that my refractor objective is collimated correctly.

I set up a Flat Panel and start to take flats (Lum filter @ 2 seconds). I then run this throught CCD inspector Flat Analyser. This I believe shows the pattern of brightness across the whole image. So given everything is "true" the image should be symmetrical. However there are two crosses on the screen, one that (I think) shows the true centre and the other where the image train is pointing. My idea is that these two crosses should be on top of one another. So this is what I did by tweaking the OTA, and now the illumination across the filed is near perfect. I know this is not a star test that CCD can do, but is it a good stab as suggesting the tilt is now "minimal" as the light illumination on the chip is even. (I probably have some technicalities wrong here, but I hope you get the idea).

Thanks

Alistair.

[Oddsocks]

Not really that accurate Alistair, though it will give you a rough idea of what is going on. 

The light distribution of a flat shows the combined light distribution of the prime optic and the scattered light from within the optical tube.

The scattered light from the anodised surfaces of all the different components in the image chain will not be symmetrical, the metal components will not be manufactured with anything like the precision of the prime optic and the depth of the anodising will not be consistent across even one component, the scattered light field will not be even and can lead you to over compensate the tilt adjustment.

You can have a perfectly collimated and tilt-free system but it may still not display a perfectly flat field.

To make matters worse after the anodising process is complete a wax is often used to seal the anodised surface after the black dye is applied and this will add to the uneven reflective surface.

You may find this document interesting reading re flats:

http://diffractionlimited.com/flat-fields-stray-light-amateur-telescopes/

 

[AlistairW]

Interesting read - lot more in this than I thought. I am guessing there should be a reasonable degree of uniformity though.

So having the "X" and the "+" symbols together is no proof of tilt correction.

Just wish there was a test I could do in the day time to look at any tilt in the system.

[Oddsocks]

I use CCD inspector quite extensively for setting up my fast Newt. 

The curvature and 3D test using a rich star field, combining at least half a dozen images averaged together, is a very accurate determination of tilt and curvature since it uses point sources that are unaffected by background gradients. I would not rely on the flat field analysis to determine tilt.

I have never tried this myself but you may be able to test for tilt using a group of ball bearings glued to black card and placed out somewhere in the sun. Provided you can place the card far enough away to focus on then a group of twenty or so regularly arranged ball bearings should make a pretty good synthetic star field on a sunny day.

You can do a basic daytime test to assure the alignment of lens cell, focuser and optical tube. It is a worthwhile check for any imaging system since you know the basic optical assembly is collimated before beginning to hang the imaging equipment of the back, any tilt then seen is entirely outside the OTA. All you need is a couple of pieces of wood and a collimation laser.

Details in this post: