Resolved STX-16803 bias and dark problem

Hi,

I have STX-16803 (purchased on February 2011, Serial: X1100119) which in the past four years was operated only a few times a year and only taking light images and it worked fine. It still works fine when taking light images. However, when taking bias or a dark (even just 1 sec) the images are nearly saturated with average count level around 47000. When taking 1 sec light image in the dark (i.e. the CCD is covered) I get the "normal" bias count of around 1600. I operate the CCD in binned 3x3 mode, but this also happens when working with unbinned.
Last time I recorded a bias and dark images is from 4 years ago and then there was no problem and their counts level were around 1600 as was the whole 7 years before.

I placed a few fits images to demonstrate the problem in http://wise-obs.tau.ac.il/~shai/tmp/bias_problem/ and their names are self-explanatory.

Is there a way to solve this bias and dark problem?

Thanks,
Shai

 

The shutter is open and close fully. No problem is seen with the shutter.
The problem is with the bias and dark and when these are taken the shutter is fully closed, but still they have a very close to saturation counts (around 47000).

 

The bias_20180627.fit and dark_20180627.fit were taken 4 years ago on the telescope and I show them for comparison of the bias and dark levels we had in the past.
All other inages were taken a few days ago in daylight (actually office light) in my office.
The inages *_with_cover.fit were taken with the metal cap screwed on the CCD.
The inages *_with_partial_cover.fit were taken with the cap a bit removed from the CCD so to let some light enter when taking an image. This is demonstrated between the images light_with_cover.fit and light_with_partial_cover.fit where the first one shows a very low count level which is basically equal to the historical bias level (whcih means that with the cap no light enters the CCD) and the later shows a high count level but not saturated, in order to demonstrate that the CCD responds to light.
When taking the bias and dark images the shutter is closed, but otherwise they are taken in the same conditions as the light images.
I did suspect light leaks so I also tested the CCD under a cover with no light around it and got the same results. Also, if there are light leaks they should have manifest themselves also in the light*.fit images since all images were taken in the same conditions.
As a layman, I would suspect that when taking the bias and dark some internal light is turned on in the camera, or that when taking these there is some problem with the electronics, but it does not make sense since when taking a light image the problem does not happen.
BTW. the recent images were taken with MaximDL V6.21 and the old images with V5.24 (but I guess this is already evident from the images' header).

 

No changes that I am aware of.
Actually these images were taken on a computer which is a new installation of the W10 system and all the software. Since I suspected this new instillation might be the cause, I took about 10 days ago the STX16803 to the observatory and checked it there on the computer in which I used for the past 4 years (which also has MaximDL V6.21) and also there got the same results.
As I mentioned above, I did not take biases and darks in the past 4 years, only light images, so I cannot tell when the problem started.

I do suppose to have at the observatory the computer from 4 years ago on which I took the old bias and dark. It is an old WindowsXP computer which was not touched over the past 4 years and still have the MaximDL V5.24. Since I am suppose to be at the observatory on Thursday I will take the STX16803 with me and try to take images on that old computer (if I will be able to turn it on . Too bad I did not think about this when I was at the observatory 10 days ago, but better late then never.

 

Thanks for the very good suggestions.

The good news is that with CCDOps I am able to take good light and dark images. The bad news is that it seems MaximDL is doing something bad to the dark images, and also to the dark images taken with CCDOps, as I explain below.

The SBIG Driver Checker that I had was V4.03 and with all the updated current drivers. I installed SBIG Driver Checker V4.05 and still it says that all my drivers are already current and there is no need to update them. Here is a screen shot of the current Driver Checker:


For the firmware, the SBIG Driver Checker says there is one driver that needs to be update:

and it says that I should not updated unless instructed by you guys, so is your recommendation to update the firmware?

Then I installed CCDOps (V5.66) and took some light and dark images and at first all images came out correctly as they should be. However, I then used MaximDL and I still get these bad bias and dark as before with MaximDL. So I tried again CCDOps and in this second time the dark came saturated as in MaximDL.
After several different attempts I found that if I power the CCD and then using CCDOps the dark images are coming OK. If I then close CCDOps and open MaximDL and just connect to the camera (not even taking an image), and then I disconnect the camera from MaximDL and start CCDOps, I get with CCDOps bad dark images which are saturated. The only way to then get good dark images with CCDOps is to power cycle the camera and then to take images with CCDOps, which are coming just fine.
This is an odd situation which I do not know what to make of it.
Seems like MaximDL, when connecting to the camera, is setting something in it that is causing these bad saturated darks.

Also, I checked for moisture: When I cool the camera condensation is seen inside, i.e., there is moisture there. What should be done to repair this?
(I should note that when the camera is not cool everything looks normal with no moisture in the chamber, but this is obvious that if there is moisture it will show itself when the camera is cold).

 

I am using MaximDL V6.21

 

attached

 

Thanks, I think updating the firmware will be straight forward using the SBIG driver checker.
Though, I am planing to be tomorrow at he observatory and will first do the test with the old computer on which that CCD worked fine 4 years ago.
Later on I will update the firmware and will check the camera with the new computer and software.

 

I tested the CCD with my old computer, on which it was working 4 years ago, and the CCD worked fine with the MaximDL 5.24 that is on that old computer. The light, dark, and bias images were all fine as was four years ago without showing the problem I have now on the new computer with MaximDL 6.21.

I then checked again the CCD on the new computer and got the same problem as reported above.

Next step was to upgrade the firmware as suggested above.
So now firmware is updated to stxga.bf2 version 2.54.
I then exited the programs and turned off the camera and back on.
Now all images are coming out correctly and the problem was solved:
On the new computer with MaximDL 6.21 the STX16803 is working fine as well as with the CCDOps.
I also checked the CCD with the new firmware on the old computer with MaximDL 5.24 and it is still working fine there, I guess this means that the firmware is backward compatible...?

Thank you for the help and for guiding me through this.

Next step is to recharge the chamber desiccant to get rid of the moisture. I will do that in the next several days and then hopefully the CCD will be ready to be mounted on our new 80cm telescope.