In a previous version of Ekos, the alignment step after a meridian flip was lost. The good news, in the latest version it is back again. The bad news, what now also starts automatically is guiding. I am imaging unguided (10micron mount) so the systems keeps hanging in the calibration step. There is no way to get out of this anymore. Pressing the stop button on the guiding page, stopping and restarting the sequence, resetting the sequence, none of it works, as soon as I start the sequence, it starts aligning and then guiding again. I had to close KStars and re-open again to get back to imaging.
Is there anyway I can turn off the guiding after a meridian flip?

so autoguiding was not used at all in Ekos up to this point? i.e. Ekos would trigger autoguiding even when no autoguiding was going on?

Correct

I‘ll check it asap. Do you have the logs at hand, please?

Wolfgang

Thank you Wolfgang.
Log is in the linked file: Log file
sorry, I had verbose logging still on (INDI and CCD), so the file is pretty big.
You'll see the meridian flip taking place around 23:27.
And in line 3566722 it moves directly from alignment to guiding.

Hope you can find the issue, and thanks for looking into this.

Willem Jan.

Hi Willem Jan,
could it be that you had non-guiding dithering on first, but turned it off at a later stage? I found in the log that you had a dithering error at 20:28:17, but after this no dithering happened until the meridian flip started at 23:27:48.

If this is the case, then we really have an edge case here, since after the dithering error at 20:28:17 the guiding state remains in state "Dithering error".

When a meridian flip is detected to be necessary EKOS evaluates whether guiding is on at this stage. And this is not that easy, since at this stage it could be that some guiding error stopped guiding. If the non-guiding dithering option is set at that point of time, guiding is considered to be off.

The mean thing in your case: the previous dithering created a guiding state that leads to the conclusion that guiding should run after the meridian flip. If you then turn the non-guiding dither option off, this last guiding state is taken as an indication that guiding should run after the meridian flip - leading to the problem that you observed.

But this explanation relies on the hypothesis that you changed the non-guiding dither option from ON to OFF during the session.

@Jasem: maybe we should consider clearing the guiding state when this option is turned off to avoid such a situation.

HTH
Wolfgang

Hi Wolfgang,

Yes, exactly that has happened. Early in the evening the system gave a dithering error and I did not have much time to troubleshoot, so I turned the non-guiding dither off during the session. So what you're saying is that all those hours until the meridian flip it had kept the state of 'Dithering error' instead of 'off' and still assumed it was doing some kind of guiding. So after the flip it assumed guiding needed to happen.

Thank you for looking into it and glad you found the culprit.

As for solutions, you have the best insight of course. But could one option be to add a few toggle switches for what you want to happen after meridian flip (e.g. alignment, focus, guiding)? Similar to the scheduler, where there is something similar.

Btw, as we're talking meridian flip settings... If you change the degree value in 'Flip if HA> ... deg', it only adjusts the time to flip if you toggle the flip on/off. Nothing major, but something to stay alert to. More intuitive behaviour would be that the time changes as soon as you change the degrees.

Thanks again for looking into this Wolfgang,

BR, WJ

Hi Willem Jan,
good to hear that my intuition was right. I think the solution as I mentioned is quite straight forward so we will fix it with the next release. The current one is already in the final stages so it will take 2 months until it will be fixed in a stable version.

Regarding your idea with the toggles, that sounds reasonable. I have to think about it whether this can at least be visualized. Currently, we are in a refactoring mode and are very reluctant touching the scheduler until the refactoring is through.

And regarding the limit: it's necessary to leave the edit field, you don't have to flip the MF enabling. At least it should...

All the best
Wolfgang

Hi Wolfgang,
That sounds fantastic. Thanks for taking care of this. Now that we know this, it's just something to be alert to until it's fixed. I guess the majority will be guiding anyway.
And for the limit, I just checked with simulators on a Mac at hand, and it showed the behaviour like you mention. My impression on the scope-computer is it's different, but will check next time. Not a biggie.

Best regards, and clear skies!

Willem Jan.