Resolved Selecting Wrong Filter

Discussion in 'STX and STXL Series Cameras' started by Mark de Regt, Aug 12, 2020.

  1. Doug

    Doug Staff Member

    Joined:
    Sep 25, 2014
    Messages:
    9,934
    If the camera is working correctly then the USB cable is not the issue.

    I'd look at the I2C cable between the filter wheel and camera.
     
  2. Mark de Regt

    Mark de Regt Cyanogen Customer

    Joined:
    Nov 19, 2014
    Messages:
    141
    We swapped the cable, and I imaged last night, but there was too much smoke to really tell. It does look like the correct filters are being selected, but it's hard to tell because of the attenuated signal.

    Also, the guiding continues to appear good.

    I'm optimistic, but I will await a truly clear night, with good seeing, to be sure.

    Mark
     
  3. Doug

    Doug Staff Member

    Joined:
    Sep 25, 2014
    Messages:
    9,934
    Okay sounds good. Hope they get the fires under control soon. Let us know if you have any further issues.
     
  4. Mark de Regt

    Mark de Regt Cyanogen Customer

    Joined:
    Nov 19, 2014
    Messages:
    141
    We've had a couple of consecutive nights with very little smoke, and better seeing.

    I am satisfied that the labeled filter was the filter actually used in every image both nights. So I think that changing the wire solved that issue.

    I also am getting fine guiding. I don't know why, for a while, the guide star took on all sorts of odd (and huge) shapes (sometimes even spinning alarmingly), but it's generally small now (the same guide star), and the guiding seems to be as good as the seeing allows.

    I think we're done here. If something crops up again, I'll come back, but otherwise, I'll leave it here.

    Thanks for your help!

    Mark
     
  5. Colin Haig

    Colin Haig Staff Member

    Joined:
    Oct 27, 2014
    Messages:
    7,380
    Location:
    Earth
    We're happy to hear this mark. Hopefully the wildfires get under control soon, or some heavy rain (one night during full moon so no time is lost). I'll mark it as resolved, but will leave it open in case you have any other info to add later.
     
  6. Mark de Regt

    Mark de Regt Cyanogen Customer

    Joined:
    Nov 19, 2014
    Messages:
    141
    Darn! It took six images, ostensibly with the blue filter, early this morning after the moon went down. Three of them were with the blue filter, but the last three clearly were with the luminance filter.
     
  7. Doug

    Doug Staff Member

    Joined:
    Sep 25, 2014
    Messages:
    9,934
    Hang on... that doesn’t sound like a hardware problem. If the wheel isn’t commanded to turn it’s just going to sit there.

    Something told the filter wheel to change slots. Sounds like a software or configuration problem.
     
  8. Mark de Regt

    Mark de Regt Cyanogen Customer

    Joined:
    Nov 19, 2014
    Messages:
    141
    I have attached the relevant ACP log, and the relevant Maxim log. The ACP log clearly says that it's selecting the B filter, but, equally clearly, the L filter is being used for those images.

    FYI, in interpreting the Maxim log, the L filter is slot 0, then, in order, R, G, B, Ha, OIII and SII.

    The only images taken last night were after the moon was well set. It took one "plan" of three 20-minute images with the blue filter, which are correct, starting at around 3am. Then, immediately following (starting around 4am), another identical plan, but it used the L filter (easy to tell the difference when looking at the actual images).

    This only happens when starting a new "plan," ACP-speak for an hour block of imaging a particular target with a particular filter. It switches to the L filter for the pointing exposure, then doesn't actually switch to the appropriate filter (although the log says it did). And it only happens intermittently.

    If it's a software problem, I have no idea whether it's Maxim or ACP, obviously.
     

    Attached Files:

    Last edited: Aug 30, 2020
  9. Doug

    Doug Staff Member

    Joined:
    Sep 25, 2014
    Messages:
    9,934
    Okay, that's progress. The question is, then, why are you seeing this problem, but no one else is reporting it? Something is different about your setup.

    I think we may want @Bob Denny to wade in on this. We need to figure out how to isolate the problem.
     
  10. Mark de Regt

    Mark de Regt Cyanogen Customer

    Joined:
    Nov 19, 2014
    Messages:
    141
    Ok; I posted over on the ACP forum. Hopefully, Bob will chime in at some point. If you have a back-channel way to communicate with him, that might help.

    Mark
     
  11. Bob Denny

    Bob Denny Cyanogen Customer

    Joined:
    Oct 12, 2014
    Messages:
    1,115
    Location:
    DC-3 Dreams, SP, Mesa, Arizona +1 480 396 9700
    OK, for reference, when ACP starts a block of images, it first takes a pointing exposure for plate solving. It bins this as high as practical (for sensitivity) and makes it as short as possible (for efficiency) and also selecta a filter you choose for this purpose, for maximum light (L, V, R), allowing the shortest exposure, and thus maximizing efficiency. Once it has acquired and solved this image, adjusted the mount as needed, it switches the filter over to the one that you specify in your observing request to use for the "real" images. So from your description the first block successfully made the switch from L to B but the later block failed to make that filter switch.

    Looking at the logs you posted, from the ACP Run Log:

    11:02:30 Switching from B to L filter for pointing exposure
    ... other stuff you can see in the log ...
    11:03:34 Switching from L to B filter for imaging


    And from the MaxIm log you posted:

    11:02:31*4 Filter wheel moving to position 0
    ...
    11:03:36*4 Filter wheel moving to position 3


    It looks like MaxIm received the filter change from ACP. I'm not sure where to go from on my end. Plus if this is intermittent, I can confidently say that if you see those "Switching from ... to..." messages in the ACP log, the filter is being changed.
     
  12. Bob Denny

    Bob Denny Cyanogen Customer

    Joined:
    Oct 12, 2014
    Messages:
    1,115
    Location:
    DC-3 Dreams, SP, Mesa, Arizona +1 480 396 9700
    I should mention that I'm seeing a lot of these

    10:43:44#5 Error - Error in plugin callback (400) - Could not activate relay - already active

    in the MaxIm Log. Could the controller be getting jammed up or ??? I have no idea what they mean.
     
  13. Mark de Regt

    Mark de Regt Cyanogen Customer

    Joined:
    Nov 19, 2014
    Messages:
    141
    Thanks, Bob.

    And just to clarify:

    1. It is always the L filter being used instead of the correct filter.
    2. It has happened with all the other filters at one time or another (the actual images taken with the L filter, despite saying they were taking with another filter).
    3. It is very intermittent (less than half the time).
    4. It always happens only for all images in a particular plan; the pointing image is taken with the L filter, the log says it's switching to the other filter, but the images for that plan all are clearly taken with the L filter.
     
  14. Colin Haig

    Colin Haig Staff Member

    Joined:
    Oct 27, 2014
    Messages:
    7,380
    Location:
    Earth
    Here's a thought.
    1. What if you don't use ACP and use the Autosave function in MaxIm to run a sequence of exposures with various filters?
    2. In MaxIm, Camera Control Window, Expose tab, Options, turn on Exposure Delay. Give it a ridiculous time like 20 seconds. Try again. Then run ACP and see if you get a different behaviour with an exposure delay in place.
    Am wondering if there is some timeout happening.
     
  15. Colin Haig

    Colin Haig Staff Member

    Joined:
    Oct 27, 2014
    Messages:
    7,380
    Location:
    Earth
    Mark, it would also be good to get:
    a) both ACP logs and the MaxIm log, so I could scan through them and see if there is anything obvious.
    b) Zip up your MaxIm Settings folder including all subdirectories and Upload here. That way I can do a fast sanity check.
     
  16. Mark de Regt

    Mark de Regt Cyanogen Customer

    Joined:
    Nov 19, 2014
    Messages:
    141
    Hi Colin,

    I attached the relevant ACP and Maxim logs to my post above from noon Saturday.

    I have attached the Maxim settings to this post.

    I'll play with Maxim as you suggested, with the caveat that, since this is an intermittent issue, if the results are negative, that doesn't prove anything. I'll point the scope at an emission nebula (which should make it easy to differentiate images taken with the L filter from all others, especially with a full moon up), and alternate images with the L filter with others.

    Also, FWIW, I do not detect any wrong filter choice in the flats--When integrating the flats, all the masters have almost no rejection, and all are slightly different (showing differing amounts of dust, and slightly different vignetting).

    Mark
     

    Attached Files:

    Last edited: Sep 1, 2020
  17. Bob Denny

    Bob Denny Cyanogen Customer

    Joined:
    Oct 12, 2014
    Messages:
    1,115
    Location:
    DC-3 Dreams, SP, Mesa, Arizona +1 480 396 9700
    Addressing the points:
    The L filter is always used for the pointing updates at the start of a run. Switching away from that is what's failing.
    After the pointing update (with the L filter) it diesn't matter what other filter is selected by ACP for your final/data images, the filter remains on L. It failed to switch to the other filter.
    This points to hardware problems.
    Yes, the pointing update is taken with L, and then after that ACP switches the filter and then takes all of the images for that filter. If your Plan asked for images in multiple filters maybe one of the filter switches would succeed. But you are using the Scheduler, and it happens that the ACP runs are always for a single color (that's how the Astro Imaging plan forms break up the work).

    My prediction: after this extensive discussion and software-based analysis, this is going to be a sticky filter wheel.
     
  18. Doug

    Doug Staff Member

    Joined:
    Sep 25, 2014
    Messages:
    9,934
    Thanks Bob, that really clears things up.

    So the question is, why isn't the wheel turning reliably? Either it's an intermittent wiring problem, or it's mechanically sticking.

    First thing to try is CCDOPS and it's Filter menu / CFW Exerciser command. This will let you run a stress test while observing whether the wheel operates correctly.
     
  19. Mark de Regt

    Mark de Regt Cyanogen Customer

    Joined:
    Nov 19, 2014
    Messages:
    141
    My thanks to both of you.

    I am inclined to agree with Bob.

    Last night, I spent some time "imaging" (meaning taking short images) only with TheSky X and Maxim, alternating between the L filter and each of the other filters. It always made the switch.

    Then I overrode the moon-avoid parameters in some ACP plans, and told it to image (no L images), and it did all of them fine.

    I'll ask the guys at SRO to run the CFW Exerciser command, and hear the filter wheel turning.

    Is there something they can do with the filter wheel (lubricate?) to lessen the likelihood of this happening?

    Mark
     
  20. Colin Haig

    Colin Haig Staff Member

    Joined:
    Oct 27, 2014
    Messages:
    7,380
    Location:
    Earth
    It may error out if there is a problem.
    What's the serial number on the FW if you have that?
    Am going to loop in @Bill in case he has any ideas, or if there is anything about this wheel that might be useful to know.
     

Share This Page