STL-4020 Strange lines

Discussion in 'General' started by Tolga Gumusayak, Nov 26, 2014.

  1. Tolga Gumusayak

    Tolga Gumusayak Standard User

    Joined:
    Nov 26, 2014
    Messages:
    28
    I was wondering if anyone knows what these 6 lines on the top left of the flat image. These lines are also visible in Lights and won't completely go away with processing.


    Screenshot (39).png
     
  2. Jan Soldan

    Jan Soldan Cyanogen Customer

    Joined:
    Oct 11, 2014
    Messages:
    239
    Location:
    Czech Republic
    Did you notice those lines also prevously or just now ?
    Could you send me (or upload here) one or two images in the FITS or SBIG file format ?
    The best solution would be to create a log file using SbigDriverChecker64 -> Debug Log button -> All button,
    then grab one flat only. You could send me this log file as an attachment for inspection.
    The image size of this camera should be 2048 x 2048 pixels. It would be possible to redefine an image size
    of this camera by removing those 6 lines on the driver level, but _only_ if related to all STL cameras with
    this CCD chip.
    Jan
     
  3. Tolga Gumusayak

    Tolga Gumusayak Standard User

    Joined:
    Nov 26, 2014
    Messages:
    28
    Hi

    Thank you for your help. I am going to email you the fits files. These were taken with no OTA. The only thing in the optical path is sensor glass. If you redifen the image size, wouldn't that crop the sensor?
     
  4. Jan Soldan

    Jan Soldan Cyanogen Customer

    Joined:
    Oct 11, 2014
    Messages:
    239
    Location:
    Czech Republic
    probably not, because imaging area of the chip is a bit larger and the subimage 2k x 2k can be shifted vertically, either up above 6 pixels or down up to 8 pixels
    without cropping. The second possibility would be to crop image height, i.e. the height of the chip would be: 2048 - 6 = 2042 rows. I have checked the Kodak KAI 4020
    specification and SBIG used the right values. I could easily generate a new experimental driver for you just to see if this helps or not. I will also check my
    STL-6303 if I see something similar, although the chip is a bit different...
    Just see, your images arrived.
    Jan
     
  5. Tolga Gumusayak

    Tolga Gumusayak Standard User

    Joined:
    Nov 26, 2014
    Messages:
    28
    I mean if you think that would help by all means lets try the exp. driver. At least it would tell us if it's software or a defect in the chip
     
  6. Tolga Gumusayak

    Tolga Gumusayak Standard User

    Joined:
    Nov 26, 2014
    Messages:
    28
    I am not an engineer but these lines look like they extend much more that 6 pixels.
     
    Last edited: Nov 27, 2014
  7. Tolga Gumusayak

    Tolga Gumusayak Standard User

    Joined:
    Nov 26, 2014
    Messages:
    28
    Is there another way to send files. They keep getting bumped back. maybe they are too large? this forum wouldn't take them either
     
  8. Jan Soldan

    Jan Soldan Cyanogen Customer

    Joined:
    Oct 11, 2014
    Messages:
    239
    Location:
    Czech Republic
    I probably badly read your message. I supposed you were talking about the first 6 top rows of your image, which looked dark, but you talked about 6 brighter
    beams comming from the left side. All your images are all right from the point of view image size definition. I also see you have a STL-4020 color camera.

    There are huge arcs visible on all your flat, you have sent me. Maybe, those arcs and beams are due to the bad light box construction or badly centered optics
    of your scope and/or small diagonal mirror, if you use a Newtonian scope. Difficult to say. In any case these beams are not produced by the camera itself.

    Jan
     

    Attached Files:

    • arc.png
      arc.png
      File size:
      442.7 KB
      Views:
      5
  9. Doug

    Doug Staff Member

    Joined:
    Sep 25, 2014
    Messages:
    6,887
    I've asked around SBIG, and learned that dark lines like that can be caused by shadows from bond wires. These are the tiny little wires that connect the CCD chip itself to its package. It's something intrinsic to semiconductor technology.

    The fact that the shadows are visible suggests some light coming from the side... perhaps improving baffling would reduce the effect.

    I remember having some really nasty reflection problems with my very first CCD camera (remember the PC-Lynx?), and it took some investigating because the problem was in the infrared and you couldn't see it visually. It turned to be light reflecting off anodized aluminum in the focuser tube - most black anodizing is useless in infrared. I made a little baffle out of blackened cardboard and put it right in the nosepiece of the camera, and the problem was gone. Maybe something like that could help in this situation.
     
  10. Tolga Gumusayak

    Tolga Gumusayak Standard User

    Joined:
    Nov 26, 2014
    Messages:
    28
    The problem is the flat was taken with no ota. So if there's a reflection it's from the camera it's self. But they show up in light frames too. When I noticed them in the lights I kept taking off glass until I got down to the camera and they were still there. They are not in the darks. So that suggests it is a reflection and not noise. Can the bond wires be coated with some anti reflective paint.
     
  11. Doug

    Doug Staff Member

    Joined:
    Sep 25, 2014
    Messages:
    6,887
    No, they are inside the chip; they're behind the sensor's window. And it's not a reflection off the wires, it's a reflection passing through the wires and making a shadow; that's why the lines are dark.

    This is why I suggested a baffle right in front of the camera, to minimize any stray light entering the camera at an angle.

    Also flocking paper or flat black paint on the inside of the focuser tubing is always a good idea (flat black paint is actually black in infrared; most anodizing is actually clear in IR).
     
  12. Tolga Gumusayak

    Tolga Gumusayak Standard User

    Joined:
    Nov 26, 2014
    Messages:
    28
    Can I send the camera in for you guys to do what ever has to be done to get rid of the lines?
     
  13. Doug

    Doug Staff Member

    Joined:
    Sep 25, 2014
    Messages:
    6,887
    About the only thing we could do is replace the sensor, and that is not an inexpensive option! Also there is no guarantee that the next sensor wouldn't have similar lines; it just depends on how the bond wires happen to lie when the chip is manufactured.

    Other than that, your best approach would be to add some baffling in front of the camera. We can't do that for you.
     
  14. Tolga Gumusayak

    Tolga Gumusayak Standard User

    Joined:
    Nov 26, 2014
    Messages:
    28
    Is this an issue with this preticular sensor? This has to be out of ordinary. Should I contact Kodak to see if they would see this as a problem and provide a replacement?
     
  15. Doug

    Doug Staff Member

    Joined:
    Sep 25, 2014
    Messages:
    6,887
    Kodak (On Semiconductor now) would simply tell you that the sensor meets their specifications.

    Try the baffling.
     

Share This Page