Streaking in images, is this a polar alignment issue? PEC? Something else entirely?

Discussion in 'Specialty Cameras SG-4, AllSky, ST-I' started by Maneesh Yadav, Mar 31, 2017.

  1. Maneesh Yadav

    Maneesh Yadav Cyanogen Customer

    Joined:
    Jul 12, 2016
    Messages:
    73
    Sigh, it never ends...

    On my never ending quest to get my friend up and running, after getting the autofocuser working we tried 2min exposures on the rosette nebula hoping for success but only to see that we still have some sort of tracking issue.

    The setup:
    Astrophysics 6 inch refractor F-9 focal length
    AP 1200 mount
    SBIG 402 ME camera
    DF-2 focuser
    SG -4 autoguider

    AP 1200 was set to sideral time with the hand control, we did not use periodic error correction.

    We used the astrophysics right angle polar alignment scope, checking the position of polaris at 8 and then 5 hours later and it seemed to be the correct location according to the astrophysics iphone app.

    I've attached a representative FITS files, the rosette nebula is without autoguiding and the ngc 2903 image is with autoguiding. The "streaks" seems to be in conjugate directions. Does anyone have a good guess as to what might be wrong? It doesn't feel like polar alignment since the alignment appears to be good and I think the relatively modest exposure times (2min) are long enough to worry about PEC on this mount. But I have no other guesses after that. Would love any advice that any of you have.

    P.S. I've also attached two autoguider log files (it isn't clear to me why they appear to be in different formats...this is something to do with how my friend captured them) in case that makes anything obvious.
     

    Attached Files:

  2. Colin Haig

    Colin Haig Staff Member

    Joined:
    Oct 27, 2014
    Messages:
    3,847
    Hi Maneesh, I'm just another customer, and decided to have a look.
    The Rosette image looks like the mount jumped a bit, after perhaps 2/3 of the exposure. Any chance the cables snagged on something or were dragging? You say it wasn't guided - the brightest part of the stars are fairly round, and then there is a jump and the next part is also fairly round. Proper PEC training might take out the jumpiness, if you can be sure that there was no mechanical problem. If you have PEC, and know it is properly set up, turn it on.

    Upon examining the FITS files, there is a lot of information that is not set up - e.g. Lat/Lon of the observing site, the optical parameters of the instrument (Focal Length, etc.) You might want to fix this, as it will help with analysis.

    I also analyzed the image quickly with PinPoint, and it looks like the camera is at position angle 82 degrees. You might consider rotating the camera to be at 0 degrees or 90 degrees - so that movement across the pixel array corresponds to the motion of the mount.
    When looking at the log, it looks like the guider is rotated 30 degrees.
    Again, you might want to rotate that so that the pixels are aligned with the mount motion.

    Also, it looks like SG4Log is the calibration part, and the SG4Log1 is the actual guiding. The calibration only seemed to move the mount small amounts in +X and +Y. Not sure why.
    During the guiding, it looks like the guider did very little for the first while, and then started to trigger relays.
    I don't really know enough about the SG4's behaviour to give useful advice. You might want to adjust settings in the calibration phase to do bigger, longer moves.

    The NGC2903 image looks like guiding didn't work in one direction. e.g. it could be a fairly good polar alignment.
    Did you guys calibrate the guiding function properly? e.g. do you know if the guider was able to move the mount sufficiently in each axis direction to determine guide times?
    Did you set the AP1200 for 0.5x guide rate or better?

    Also, I see the software is very out of date, MaxIm 5.12 - you might want to upgrade to at least 5.24, and maybe go all the way to 6.13 if you can afford to upgrade.

    After dealing with the other items above, I suggest you make a few 10-20 minute exposures, if your skies are dark enough. 1 or 2 minutes aren't showing enough of what is going on.
    Then post those, and maybe we can figure out what is going on.
     
  3. Maneesh Yadav

    Maneesh Yadav Cyanogen Customer

    Joined:
    Jul 12, 2016
    Messages:
    73
    Great advice thanks!
    Will follow through on these and get some more exposures (will be a few weeks as I am away).

    Only reason that I haven't upgraded the MaximDL version is that I really want to get stuff to the point of "working" ; will happily upgrade once we eliminate all these sources of error.
     
  4. Maneesh Yadav

    Maneesh Yadav Cyanogen Customer

    Joined:
    Jul 12, 2016
    Messages:
    73
    Just a quick update.
    I think we were able to address some of the Colin's suggestions: site info etc., certainly weren't any cables tugging (picture of cords attached). I got a hold of PemPro to check the polar alignment and PEC; I'll be posting my PEC data to the ccdware forums to cross check, but (assuming I am interpreting the graphs correctly) I was able to see a clear improvement in tracking after improving the polar alignment and applying the PEC correction. The polar alignment adjustments were significant.

    Still not sure if we are tracking credibly (autoguiding is turned off), I've attached images today for 2 and 4 minute exposures of M3. The 2 min exposure looks ok to my eyes and the 4 min exposure clearly is clearly streaking (uploaded 2 images of each). We were anxious to see what things looked like with autoguiding but the SG4 interface was, somehow, being problematic. We've gotten it to work before but today we could connect to the autoguider fine, but clicking "calibrate guider" would just freeze the SG4 interface. Tried shutting down all other programs, checked the phone cable to the mount and no obvious signs of something being wrong. No real changes to the computer either and tried following the manual to the letter. Any ideas would be welcome (just attached a picture of the interface, I swear "take image" works fine).

    Screen Shot 2017-05-04 at 12.57.29 AM.png IMG_3092.JPG
     

    Attached Files:

  5. Colin Haig

    Colin Haig Staff Member

    Joined:
    Oct 27, 2014
    Messages:
    3,847
    Hi Maneesh, glad to see you are making progress.
    I agree with you. The 2 minute exposures look quite good, and the 4minute ones are streaking.
    I hope the Diffraction folks can offer some suggestions on the SG4 issues.

    One thing that would be helpful - can you upload some FITS files from the guider? The screenshot looks like there are wavy lines in it - these are possibly caused by electrical interference or radio frequency interference (EMI/RFI).
    Do you share power between devices? It might be worth putting a few ferrite cores on the cables or changing power supplies.
     
  6. Doug

    Doug Staff Member

    Joined:
    Sep 25, 2014
    Messages:
    7,308
    A few thoughts:

    1. You have a lot of cables dangling from your setup. Dragging cables are a leading cause of bad guiding. Tie them off and route them down the axis of the telescope, rather than dangling from the end.

    2. Make sure the mounting of the guide scope is really solid. It only takes ~10 microns of flexure or slop to cause guiding errors like this. Yes that is an incredibly tiny distance.

    3. You're obviously using an SCT. Does it have a mirror focus lock capability? Mirror flop is an extremely common cause of bad guiding with SCTs.

    In short... I suspect your problem is mechanical in nature.
     
  7. Maneesh Yadav

    Maneesh Yadav Cyanogen Customer

    Joined:
    Jul 12, 2016
    Messages:
    73
    Thanks Colin and Doug.
    Will upload fit files from autoguider next chance I get.

    Just to be clear, I *could not* get the autoguider working last night, the fit files in the previous message taken unguided. Doug I presume you are suggesting that the "wavy" image quality may be due to poor mounting, which also might prevent "calibrate guider" from coming up correctly in the interface.

    Will follow advice on trying cables and check the guide guidescope.

    Thanks again!

    EDIT: Meant to add that the C-14 edge ED tube has built in mirror lock system so the mirror does not flop when it crosses the meridian
     
    Last edited: May 4, 2017
  8. Doug

    Doug Staff Member

    Joined:
    Sep 25, 2014
    Messages:
    7,308
    The wavy lines are some kind of noise, but they are very faint and won't affect SG-4 operation. Make sure your stars are reasonably in focus and try again.
     
    Maneesh Yadav likes this.
  9. Maneesh Yadav

    Maneesh Yadav Cyanogen Customer

    Joined:
    Jul 12, 2016
    Messages:
    73
    Aggh. Still so much pain.

    After obtaining what appeared to be semi-respectable unguided tracking (I believe primarily as a result of PEC with a newly installed GTOCP3) 2 days ago, I tried autoguiding today.

    One thing that worked better is that the software buttons all worked (they, somehow, didn't always work earlier even when the camera was successfully connected), I have a sneaking suspicion this was due to the general pokiness of the PC that is controlling things and running windows 7. I did some standard things to eliminate windows 7 lag and things are much snappier now.

    Now in trying to eliminate the problem step by step:

    As a control I tried 100s exposures that now I'm confident should not be problematic unguided, and they seem to look fine (I've attached a representative image, through blue filter to avoid saturation on longer exposures). This supports the notion that dragging cables etc. are not contributing to error in unguided tracking.

    - Now I believe a 100s exposure with the guider should give at least the same image if I've done things correctly. We focused and calibrated correctly (arrg I forgot to save the FIT file from guider...I won't forget next session), calibration log attached. My backlash correction is set to 0 in the ASCOM panel, which I understand to be the correct setting for automated guiding. Is there anything jumping out in the log file that seems fishy at the calibration stage?

    - I then took 100s exposure with automated guiding ("ag") and you can see that tracking is poor with a streaky double image. The guiding log file is attached.

    -The SG4 manual suggests, with good PEC, one can reduce the aggression. I tried some runs with aggressions set to 0.1, the log files looks "sparse" (lots of zeros), which makes sense, there shouldn't be a whole lot of correction going on. A 100s guided image looks ok (not included for brevity), but perhaps this isn't informative with the short time and low aggression.

    -I tried a 10min exposure unguided and guided. The 10 min unguided exposure is not as nice as the 10 min I unguided image I posted 2 days ago with some elongation. The low aggression guided image is "jumpy".

    -I also have 5 min exposures showing that guiding is decreasing the quality of the image (the 5min unguided exposure seems to look fine), left those out for brevity.

    The unguided images certainly aren't always high quality as there was small trailing some 5min exposures where there didn't seem to be very much at all in a 10min image in the images I posted yesterday (as Colin noticed). PEC seems to have helped a lot but there is some more error that I haven't isolated yet. Still, autoguiding seems to be responding too "strongly" to (what I think) are small movements as the autoguided images are never better than unguided. I must be making a very fundamental/elementary mistake somewhere.

    Should I be dropping the aggression to something miniscule (0.02?)Mn and Max move are set to 0.05 and 0.5 (per manual). Hard to think of solutions as there just aren't many knobs to turn on the SG4!
     

    Attached Files:

  10. Maneesh Yadav

    Maneesh Yadav Cyanogen Customer

    Joined:
    Jul 12, 2016
    Messages:
    73
    Ok, got a pretty clear night yesterday and redid the the focus and calibration on the SG4. Log files made much more sense, with only sparse corrections even at the default aggression. I'll post on my thread in the General forum, but I think I did ok.

    I think the direction vectors were wrong on the last try; I suspect it was due to poor focus.
     

Share This Page