End of sequence after meridian flip

[giraud]

Hi Robin,

My sequence didn't go through last night. The stop (after 1.14am) occured after the meridian flip routine was automatically triggered, a routine that has never failed so far.
I got a pop-up that was warning that the flip was not advised and required to make a choice (Yes/No) to perform it or not. As I was not in front of my computer at that time in the night, I couldn't apply any choice. Could you please investigate and identify what caused this?

Thanks for you support.

Log is attached.

Cheers,
Eric.

[admin]

Hi Eric,

interesting - I think that the message and the yes/no choice was coming from your ASCOM driver, since that SharpCap doesn't contain a message that asks you to confirm a meridian flip.

Fron SharpCap's point of view, nothing really went wrong until about 1:24 when PHD2 lost guiding and could not restart. I suspec that at that time the mount got to the meridian limit and stopped tracking, turning all the stars into streaks.

The 'MeridianFlip' subroutine seems to complete without any error between 01:14:41 and 01:18:49, but I wonder if the mount ASCOM driver was holding off any movement based on waiting the response to the Yes/No question. In that case, if it didn't produce an error code, SharpCap may not know that anything is wrong with the mount - I would need to see the full sequence code to understand more about this.

Somewhere in your mount ASCOM driver settings there may be adjustments you can make to when it will allow meridian flips without confirmation. Unfortunately though this sort of thing is often poorly documented (or undocumented), so can require trial-and-error to understand what the settings do.

cheers,

Robin

[giraud]

Hello Robin

I experienced it again, commanding a plate solve after a meridian flip as part of my routine. See picture attached, it looks like a Sharpcap pop up menu to me.
I think it is linked to "Toolbar/Menu input : SharpCap.ViewModels.Actions.SolveAndResyncAction complete" at the end of the attached log. I did press ok and it went fine, but I am not always there at night to validate this

Thanks for your support.

Cheers,
Eric

[admin]

Hi Eric,

ah yes, that is a SharpCap message! I was searching for the wrong wording.

What is happening there is that the mount is say flipped past the meridian (east side pointing west at maybe 182 azimuth) according to the mount ASCOM co-ordinates. The plate solving result comes in at a position on the other side of the meridian. Sending a sync command to the mount in those conditions is a bit of a gamble - the implementation is mount dependent, so some might decide that they need to flip back to the opposite side of pier as a result of the sync.

However, the message is not ideal when running a sequence, so the question is, what should SharpCap do in that circumstance when running a sequence?

I'm wondering if the right answer is to fail and retry if the current position of the mount is just past a flip? What are your thoughts? Did the problem happen after the flip?

cheers,

Robin

[giraud]

Hello

My MeridianFlip routine is triggered when the meridian is past of 1°.
The routine (picture attached) would first record the current position (Step 1), and then immediately after require the mount to point at that same position. The mount would then detect it has to do a meridian flip (via Ascom) as it has passed the meridian (Step 2). And only after that would the platesove be executed (Step 3), and the solution shall then be on the same side of the meridan.

From what you are writing, it looks like the flip has not been performed (Step 2)?

This routine has worked flawlessly for quite a while until when I resumed my observations. I didn't see anything out of normal in the Ascom driver input parameters.

Cheers,
Eric.

[admin]

Hi,

'Stop at 1 from meridian' will stop the previous imaging when the mount is 1 degree short of reaching the meridian - 4 minutes of time. Depending on how quickly the sequence runs from that 'stop' point to getting to call the MeridianFlip subroutine, it could happen that the attempt at meridian flip using save/goto co-ordinates would run before the mount crosses the meridian, so will not trigger ASCOM to flip. I presume the sequence then continued and the next resync picked up the co-ordinates being the wrong side of the meridian.

I would add a delay to ensure that the mount has passed the meridian before the flip subroutine is called, even if the 'stop at' responds almost instantly at 1 degree before the meridian. Even then, you could still end up in trouble depending on the direction/magnitude of the pointing error after the flip. Another pause after the flip (4 minutes of time = 1 degree of rotation) would give some safety there.

cheers,

Robin

[giraud]

I understand your proposal and that should work. Thanks.

However I am still wondering; my routine calls for a Meridian Flip subroutine when -1deg from the meridian (see attached). My understanding was that the meridian was therefore passed. I am mistaken?

Cheers.
Eric.

[admin]

Hi,

that's odd - that doesn't match the code of the sequence recorded in the logfile above :

Code: Select all

        STOP AT LOCALTIME 03:30:00
            TARGETNAME "NGC 5055 Sunflower"
            MOUNT GOTO "RA 13h 15m 48.22s | Dec 42° 01' 52.1'' "
            SAVE MOUNT COORDINATES
            STOP AT 1 FROM MERIDIAN
                GOSUB "Main Imaging Sequence #2"
            END STOP AT
            GUIDING DITHER EVERY STOP
            GOSUB MeridianFlip
            GOSUB "Main Imaging Sequence #2"
        END STOP AT

I just checked the editing and load/save of that step and all looks fine for preserving a negative number if one is set, so not quite sure what happened, but SharpCap was definitely running a sequence that requested a stop one degree before the meridian, which might explain a lot!

cheers,

Robin