Why are my images progressively drifting?

[lukebl]

Can anyone explain why my images are drifting during the course of an imaging session? My PHD graph is pretty flat, and polar alignment is good, but there is a progressive drift throughout the session. Here are the first and last images superimposed from a recent imaging session. The drift isn't specifically in RA or Dec, but something in between. If this had been one long exposure, I assume it would have resulted in star trails the distance between each pair of stars, if you get my meaning.

Presumably if it was differential flexure it would be a bit more erratic, rather than a gradual and even drift in the images? Or is this clear evidence of poor polar alignment, even though I thought it was OK, and the drift isn't specifically in RA or Dec?

[bamus]

I have got same situation (more in RA I think) and would like to know answer. My PHD graph shows no drift in either X or Y axis, but last picture is many pixels away from first one. Its annoying as final stack shows bands of noise which I this is result of this slow drift.

[themos]

How are you guiding? with an OAG or a guidescope? How far away is the guide star from your imaging center? What is the declination of the target? How good is your polar alignment, in arcminutes? I'm afraid you have to fill in all these boxes before you can hope to analyze what is going on.

[themos]

I've analysed the same problem I was having, here:

http://stargazerslounge.com/imaging-discussion/165776-autoguiding-drift-diff-flex-all.html

[bamus]

In my case I am guiding with guide scope which I am sure is not perfectly aligned with main scope, but I have been told it doesn’t matter. Polar alignment is quite good as I am using drift alignment.

[themos]

"Quite good" doesn't tell you enough to analyse. You need to measure your PA error in arcminutes before you can crunch the numbers to see if you field rotation is what is going on or whether it's down to diff flexure.

[bamus]

How you are measuring PA in arc minutes? Never done it before. For me flat DEC graph is enough, don’t need to know value.

Anyway. It cannot be field rotation as my drift is in one direction throughout whole of the picture. Not sure about flexure, focal length of imaging scope is 432mm, guide scope is 135mm. maybe there is an issue?

[lukebl]

... You need to measure your PA error in arcminutes ...

How do you measure this?

[themos]

You put a bound on PA error by various methods. One is to measure the declination drift (while not guiding, obviously). here's a calculator

Polar Alignment Error Calculator

It cannot be field rotation as my drift is in one direction throughout whole of the picture.

If you feed your frames to DeepSkyStacker it will calculate the relative angles of each frame. have you done this?

[RobH]

It's most likely differential flexure.....I'm dealing with exactly the same problem with a side by side rig at the moment. I can't find anything that may move, but PHD shows good guiding and the mount isn't anything like overstressed....it's an AP1200 and I just have 2 x 80mm fracs on it at the moment.

[lukebl]

Rob, are your images evenly and precisely drifting, like mine? I would have thought that differential flexure would be less precise and even?

[themos]

No, it can be quite even as gravity pulls one way only.

[themos]

Anyway, the way to diagnose this is to measure your field rotation in DSS. Then measure your PA error and the distance of the guide star to the imaging center and plug the numbers into a calculator to see if that could be what's giving you the measured field rotation. If you are getting field rotation your image WILL appear to drift as well as the center of rotation is your guide star which may be off the imaging frame. If you DON'T get field rotation then it's most likely DF.

[FraserClarke]

I would have thought that differential flexure would be less precise and even?

True flexure should be very precise, even and repeatable. It will only get erratic (like my spelling?) if you've got some stick-slip motion going on, e.g. from a bolt not being tight enough.

[themos]

I think I've seen DF that fades out as you approach the meridian only to gradually reappear after a meridian flip. I blamed Isaac Newton.

[steppenwolf]

I blamed Isaac Newton.

He certainly has a lot to answer for - DF and I have a long history and we still don't get on that well ......

[lukebl]

Thanks for the input, gents. Back to square one. Just when I thought I'd sorted out my PA and diff flex issues.

Curse that Newton!

[themos]

No, now you're on square two!

[RikM]

I have exactly the same happening. I am using a finderguider but I haven't cemented the finder in to the rings yet. It is still only held by the original thumbscews and tension bolt thingy. I had hoped it would be light enough to hold steady as it is but I may have to do something about it.

[lukebl]

Just a thought. My guidescope is currently on the opposite side of the OTA to the focuser on my setup, as shown in option A of this illustration. However, I have seen some setups where the guidescope is on the same side as the focuser, as in option B.

Does that have any bearing on the guiding effectiveness, as long as the balance of the scope is OK?

[themos]

You mean affect the DF? Yes, it could, depending on how good the mounting is and which way gravity is pointing while you track. My hunch is that a lot of DF is at the coupling of guide-camera-body to guide-optics.

[Daniel-K]

i have mine set up like image B and im not sure what flex is but so far never experianced image's like yours but i will be watching this for more info, hope you get sorted

[RikM]

Mine is set up like B also, but I do get movement.

[lukebl]

Thanks again for the input, gents. I think I'll try it with the guidescope in position B, tighten everything up, and see if that works.

Funnily enough, I had a look back to read my first-ever post on SGL, two and a half years ago. It was a query about differential flexure. Nothing changes!

[Star Gazer]

I don't guide when I am imaging, but I have seen the same problem in my images. Obviously in my situation the problem is not DF. Also, shoot me down if you like but the distances between consecutive stars looks way to big to be DF, but then what do I know.

To me this looks like incorrect star alignment, a image processing problem. OK may be I'm on the wrong track, but I think you should consider this before looking for a solution for DF. Run the same images through another package and see where it gets you.

Just my thoughts.