Why are my flats producing 'Raised Donuts'?

[NoelC]

When stacking recently the flats have not removed the donuts entirely.  When stretching the images they re-appear as raised donuts.


the flats have been generated by APT using the Flats Aid tool, and in the past have been pretty good.  Apart from a few revisions of APT I've not changed anything; anyone any ideas why the flats now appear to be producing the wrong result?  Master flat for above:
Tried stacking with bias and without, haven't actually taken flat darks.

Flat Aid target is to produce flats at approximately half ADU (32k).

I should point out (for those wincing in horror) that this is after splitting and cleaning the telescope optics!

[Carole]

I had something very similar happen recently Noel.  After some discussion on SGL some-one suggested it could be due to the dessicant needing changing. 

I changed the dessicant and gave the camera a chance to dry out and the problem disappeared, so might be worth trying.

I'll see if I can find the picture I posted up at the time for comparison.

Carole

[Carole]

I have linked this from SGL, hopefully it will show up here.  This is a stacked result showing what i think are similar results Noel. 

[NoelC]

Thanks Carole
That's exactly what I am seeing (as if the donuts are lit from the side).
In my case the CCD has no desiccant to change.  Not sure where to go with this.
I do notice that my flats have a gradient; assumed this was due to the OAG prism; but could be due to the flat panel (it's LED not EL).  Will try re-doing the flats.

Noel

[Carole]

I am surprised you have no dessicant, perhaps the camera needs to be placed in a dry place to dry it out.

Carole

[ApophisAstros]

What camera have you got Noel?
Roger

[NoelC]

Re-shot the flats. turning the flat panel as they went to eliminate any gradient from it. 
Result is almost exactly the same.
If I take the camera off; I will lose the alignment and be unable to create new flats.
The flats level is hovering around the 33,000adu mark (which is OK - but a little over 32,000).
Tried re-stacking using average, median and sigma 2 - all have the same artifacts.
I'm beginning to wonder if I kicked the camera or somthing like that in putting the scope away.
Roger:
I'm using an Atik One 6.0.

[ApophisAstros]

Found this in the manual "The Atik One features a CCD chamber that has been purged with high purity Argon to eliminate moisture from the sensor and to optimize cooling.", if this has the same effect as a dessicant tablet and its not as it should be then it may cause your problem, if it is i am sorry but no advice , only to contact ATIK as some companies repair things for free as with my Lunt solar scope.(hopefully i wont need to take advantage of this)
HTH
Roger

[Carole]

I had to ditch my subs Noel - get the camera dried out and start again.

I hate to say this but that camera has cause a few problems one way or another, the fact that you only have 5 filter positions being another draw back, and now this.

Atik are extremely helpful even after I had a problem with a second hand camera they repaired it for me.  Not free of charge of course. 

Carole

[NoelC]

Thanks Carole
I think I found the problem; my camera was loose!
The focuser is screwed onto M90 extenders; must have turned the focuser CCW slightly when trying to get the guide cam orthogonal to the mount.

Digging around on the forums, apparently you get very similar artifacts from changing the focus position, but the front runner is moving the camera with respect to the telescope/optical train.
https://www.cloudynights.com/topic/676521-what-are-these-ring-artifacts-poor-flat-calibration/?hl=%2Braised+%2Bdonuts#entry9643220
As the camera only progresses the filters in one direction, that was unlikely to be the cause. 
Re-took the flats again with the camera propped on a workbench (pushing it up the other way) and the results are slightly less bad.

[MarkS]

I think I found the problem; my camera was loose!

That would explain it!

Mark