ASI2600MC Duo: Diagonal gradient banding on the flats

Yuri

Dear fellow astrophotographers and the ZWO Team,

Kindly seeking a bit of your advice on a strange issue with a brand new ZWO ASI2600MC Duo camera.

Since its first light in March, this camera shows strange (although subtle) diagonal bands on the flat and light frames. Broad bands run across the image at almost perfect 45 degrees: below are an example flat frame and an example light frame as seen in ASIFitsView application. FITS images and the result of PixInsight's FlatContourPlot script are uploaded to OneDrive: link to the files.

I tried to figure out what could be causing this and performed a series of experiments. Found out that the banding did not depend on:

Light source for the flats
Camera rotation angle
Filters
Devices between the scope (Askar FRA500) and the camera. My regular imaging train is FRA500 -> ZWO CAA (M54 thread to the scope) -> ZWO EFW 5*2" (M54 thread to the CAA) -> Camera tilt plate M54 (bolted to the EFW) -> Camera (bolted to the tilt plate). Removing anything in between the camera and the scope (I leave the tilt plate on, though) does not change the state of things
Possible light leaks in the imaging train (tried to cover all thread connections with a cloth)
Exposure time for the flats
Camera gain setting (although the banding seems to be more pronounced with lower gain values -- but that may be an effect of the image stretch)
Camera cooling setting or dew heater on/off state
Power supply. Same thing happens with both mains power and a 12V DC supply
Suspected gap between the tilt plated and the camera (I have taped out that gap)

There is no such banding on dark frames. Camera window is clean with no sign of oil leaks or smudges. Calibration in PixInsight seems to remove the banding.

I used to have an ASI2600MC Pro before, and I have never encountered such an issue.

Has anyone noticed anything similar? All suggestions are highly appreciated.

Thank you.

Kind regards,

Yuri

Tech@ZWO

Yuri Would you mind sharing some fits files of different field frames with us? these maybe helpful.
Via google drive or other cloud drives. thanks.

Yuri

Thank you. It looks like the link to OneDrive folder with the images did not go through. Here it is: Link

Plus a flat and a light frame from the same imaging session.

Tech@ZWO

Yuri It looks like watermarks left on the protective glass, sensor, or filter. Have you cleaned them recently？ You can rotate them one by one to tell where the marks are easily.

Yuri

Tech@ZWO

Thank you for reply. The camera was brand new, from a sealed box. If those are watermarks, how could they appear then?

Filter is out of question, as banding does not depend on the filter presence.

I have cleaned the entire outer surface of the protective window: no result, banding on the flats look the same. Quite afraid of going further with the sensor cleaning.

Is sensor cleaning a user-level maintenance operation? Do you provide instructions how to perform it?

Yuri

Tech@ZWO

UPDATE.

Just cleaned the inner surface of the protective glass. The banding is still there, nothing changed. So the glass does not seem to be the cause.

Bare sensor looks perfectly clean.

Shall I try cleaning it with the sensor swabs for APS-C cameras?

probro1154

I'm not seeing the banding in the light you uploaded to onedrive. See it in the flat, but not the light. Messed around with fitsview's stretch trying to recreate, but couldn't.

Have never seen banding from my own Duo, but haven't been doing flats either. Here's a single light from mine (auto stretched and saved as a PNG file, but no other processing): https://imgur.com/gallery/m31-single-300sec-exposure-zwo-ff80-2600mcduo-RSXXwlM

Yuri

probro1154 , thank you for your response!

Indeed, ASIFitsView does not show banding in that light frame, but here is what PixInsight's FlatContourPlot script finds in the same frame:

probro1154

Interesting! I can duplicate with pixinsight, but running the same script on some of my lights doesn't show anything like that. Even a 30sec light, like yours. I was going to ask what gain you were using, but see 100 in the fits header. Hopefully ZWO can help you figure it out!

Yuri

probro1154

Thank you for your research. I used to own an ASI2600MC Pro and an ASI294MC Pro before, and have not seen such banding pattern with either of them.

It is only my ASI2600MC Duo who reveals this pattern from day one.

I am still hesitating to clean the sensor, waiting for a confirmation from @Tech@ZWO . After cleaning inside and outside of the protective window (with no result) yesterday, sensor has already managed to collect a host of dust bunnies.