Page 1 of 3 123 LastLast
Results 1 to 10 of 27

Thread: Gemini Flipping

Hybrid View

  1. #1

    Default Gemini Flipping

    Bob,

    While I'm pinging you... You know I'm still having issues with my flipping... You should be able to see what's happening in the logs I uploaded. What it looks like is that well before the flip point it does a "flip slew" (which doesn't flip the scope), and then tries to image which messes up the guiding directions for one, and wastes a mess of time for another. Usually I end up having to manually flip the mount and restart my plan. I'm pretty confused about this, The driver for the Gemini is supposed to know all about its flip state, so, I'm not sure why this is occurring.

    Thanks,
    -esy
    Attached Files Attached Files
    My punctuality is well known. When "The Revolution" takes place, I'll be late, and I'll be shot as a traitor.

  2. #2
    Join Date
    Oct 2005
    Location
    Mesa, AZ
    Posts
    28,502

    Default

    If this is the case, the flip points in ACP are mismatched to the actual ones in the scope. Don't assume the driver is supplying flip info. In order for ACP to use flip info the driver must implement both PierSide and DestinationPierSide. I haven't seen a driver yet that implements the latter. For some reason the authors assume it's fluff or it's too hard to do or ??? I'm guessing that the Gemini doesn't implement it, or worse, it's giving an incorrect answer. If you click the Flip Settings... button do you see a popup that says "your telescope reports its pier side"? If so, then it is getting both from the driver.
    -- Bob

  3. #3

    Default

    According to Paul, the driver supports PierSide, but not the DestinationSideOfPier method.

    -esy
    My punctuality is well known. When "The Revolution" takes place, I'll be late, and I'll be shot as a traitor.

  4. #4
    Join Date
    Oct 2005
    Location
    Mesa, AZ
    Posts
    28,502

    Default

    OK. So in ACP, just verify that when you click the Flip Settings... button (Telescope tab) you do not get a popup saying that your telescope reports its pier side.
    -- Bob

  5. #5
    Join Date
    Oct 2005
    Location
    Mesa, AZ
    Posts
    28,502

    Default

    I'l look at your logs Monday... it's really hard to tell much without being there. I'm guessing that the flip points aren't right though, and that you have been assuming that ACP is getting that info from the driver. If only these new drivers would support DestinationPierSide! It's there in the interface for a reason :-))) I think AP is going to support it - the developer misunderstood what it was but after re-reading the spec he said it was reasonably clear. If you want Paul K to get this info pronto, I hope you don't mind passing it along to him. As I said in the other thread, I'm not where I can get to my email.
    -- Bob

  6. #6

    Default

    So, I passed along this message to Paul. I think we have him convinced to support that method: he said he sent an email to Rene to get more info.

    Until that happy day comes however, I've reverted my flipconfig.txt back to the defaults (300 pre, 120 post). Last night I had a horrible time around the meridian and basically lost an hour of integration time.

    The real problem as I see it, is that ACP seems to think the mount has flipped, when in actuality it hasn't. in the logs you should be able to see this. When this happens it inverts the guider direction and all heck breaks loose. I don't understand how it could not know what side of the pier it is on when the mount definitely reports SideOfPier. Is it possible you need a var you can set at the beginning of a plan/run, then update it when you think a flip has happened, then check it against SideOfPier? I think such a scheme would preclude this from happening. of course an alternate plan would be to issue a meridian flip command when it is time to flip. Of course, the problem with that, is that there is no ASCOM method to order a flip per se, even though the mount has an actual command to direct such. Why isn't there an ASCOM method for that, by the by, would have seemed a logical method to have added to 2.0?

    -esy
    My punctuality is well known. When "The Revolution" takes place, I'll be late, and I'll be shot as a traitor.

  7. #7
    Join Date
    Oct 2005
    Location
    Mesa, AZ
    Posts
    28,502

    Default

    Yep. Without the mount reporting pier info, you absolutely positively must get the Flip Settings (ACP Preferences, Telescope Tab, Flip Settings... button) correct. Once the Flip Settings window is open, note the small hyperlink "What is this?" in the bottom left. Click it to go to the German Mount info page in ACP Help. There you will find the procedure for determining the true flip points (both east- and west-bound) for your mount. If you change anything in the Gemini handbox that changes flipping behavior, you have to go back through those tests and make the adjustments needed to properly reflect your scope's flipping behavior.
    -- Bob

  8. #8
    Join Date
    Oct 2005
    Location
    Strongsville, OH
    Posts
    1,548

    Default

    Hi Erik,

    Why isn't there an ASCOM method for that [flip], by the by, would have seemed a logical method to have added to 2.0?
    I don't think this is an ASCOM issue. To the best of my knowledge, none of the mounts implement a "flip" command. When they are commanded to slew to some RA/DEC, the mount internally decides whether it needs to flip or not. This isn't determined by ASCOM mount drivers. This is determined by the mount's firmware.

    That's only fair, isn't it? After all, the guy who makes the mount should be the one that determines the flip behavior, not some 3rd-party software developer!

    FWIW.

    Jim

  9. #9

    Default

    I'm not saying it's an ASCOM issue per se. Just seems like an oversight. It's not intended as a critique, simply curiosity.

    Gemini has a command string dedicated to flip the mount. I don't see why that shouldn't be leveraged instead of commanding a slew. I mean a flip to a location and a slew to a location are not necessarily identical, as a slew is dependent on SideOfPier and DestinationSideOfPier() and as you point out, the definitions/decisions/parameters of the mount firmware. A flip slews to the current position (presumably, if possible, and with the obvious restrictions) but looking from the other side of the pier. I view those as distinct (though obviously not orthogonal) operations, with different semantics. And now I think about it, presuming the mount to have suitable restrictions in place to govern flipping behavior when calling the flip command, Wouldn't that actually be a preferable methodology to command a flip, that way one is insured of a flip where possible, and it could (theoretically) be checked.

    The issue here I think is not one of the mount's flip semantics, but of ACP's. Now if ACP's and the ASCOM semantics are essentially one and the same (i.e. they degenerate to the mount case by commanding a slew), I don't see how that explains the behavior i'm seeing.

    My particular problem (would there was only one! ) with flipping is really revolving around what I see as ACP not recognizing that the mount has, in fact, NOT flipped. I mean, I would be totally cool if it just waited until the object was actually past the meridian before flipping and delayed the plan accordingly. I have tried even with pre-flip margin set wildly negative (up to -1200) and the results are the same, it does this little dance. Now, I certainly concur with Bob that I need to find the time to configure the flip settings more precisely (at least until Paul adds DestinationSideOfPier()). Perhaps I can write a script to determine these settings for me?

    And, perhaps I'm dense, but even ignoring the issue of DestinationSideOfPier(). Gemini L4 does report SideOfPier, hence ACP should always know which side of the pier the scope is on. Am I misunderstanding this property? I mean, my situation is that ACP thinks the scope has flipped, and has slewed to the object from the other side of the pier, but clearly that isn't the case, which should be able to be validated by the platesolve performed after the putative flip and the SideOfPier property.

    In the grand scheme, maybe this is all a tempest in a teacup, and perhaps more an educational discussion. In which case please do enlighten me.


    -esy
    My punctuality is well known. When "The Revolution" takes place, I'll be late, and I'll be shot as a traitor.

  10. #10

    Default

    And so we're all on the same page, here's some log (from just now as it happens) I am currently about 54 minutes from transit, and no flip has actually taken place:

    Code:
      (GEM must flip before next operation)
      (waiting 484 sec to pass flip limit)
      (wait ends at 01:57:26 UTC)
      (turning tracking off)
      (wait finished, resuming tracking)
      (flipping mount...)
      Start slew to M31-Mosaic-A2...
      (wait for slew to complete)
      (slew complete)
      (flip completed
      (doing post-flip pointing update...)
      Updating pointing...
      Switching from Green to Luminance filter for pointing exposure
      Focus change of 13 steps required
      (taking 10 sec. exposure, Luminance filter, binning = 2)
      Set FLI - New readout mode to 8MPPS.
      Image finished
      Plate-solve pointing image.
      272 image stars found
      879 catalog stars found
      Solved! 203 stars matched.
      Average residual is 0.36 arcsec.
      Pointing error is 31.746 arcmin @ angle 79.09
      True focal length is 77.8 cm.
      True image center (J2000): 00h 45m 30.5s   40 18' 26.43"
      Imager sky position angle is 270.6 deg.
    **SYNC ACROSS FLIP POINT FOR GERMAN MOUNT. SYNC SKIPPED!
      Re-slew to target.
      Start slew to M31-Mosaic-A2...
      (wait for slew to complete)
      (slew complete)
    And the guider directions just went flippy on me.
    And for good measure, I just commanded a manual meridian flip (still 40 minutes from transit), which went fine.

    -esy
    Last edited by Erik Young; Nov 19, 2009 at 01:16. Reason: additional informarion
    My punctuality is well known. When "The Revolution" takes place, I'll be late, and I'll be shot as a traitor.

 

 

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Similar Threads

  1. Help Requested with Gemini Mount Flipping
    By Robert Brewington in forum Hardware/Software/Driver Topics Not Directly Related to Our Software
    Replies: 4
    Last Post: Mar 30, 2011, 15:46
  2. Gemini Flipping Problems - Again
    By Joao Luis Gregorio in forum Hardware/Software/Driver Topics Not Directly Related to Our Software
    Replies: 6
    Last Post: Feb 26, 2010, 14:14

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •