Hello!

A few weeks ago I began controlling my rig with a raspberry pi4 configured with Astroberry/Kstars. I almost have everything working perfectly, but there are a few issues I have not been able to figure out.

The most pressing issue has to do with the meridian flip. I have the auto meridian flip box checked, and it is set to flip once the hour angle passes 0 degrees. When the flip first initiates, the dialog box says it is flipping. However, after a few seconds it says the flip failed and that it will try again in 4 minutes. I think it slews a little because after the failure ekos begins the plate solving procedure. This cycle repeats itself until I abort the imaging plan. I've tried manually reslewing to the target in kstars, but it does not flip. Only until I park the scope and completely restart ekos does it flip. So far I have tried:

updating everything on the pi
enabling daylight savings on the hand controller
updating the location on the hand controller
disabling auto meridian flip on the hand controller

For clarification, my mount is a skywatcher EQ35, and it is connected to the pi via a serial port connected to the hand controller, which is set to pc direct mode. I ran this same setup but with my PC and different imaging software(APT and then NINA) for over 2 years with absolutely no problems, so I don't suspect the problem lies in hand controller settings on the mount.

I'm also trying to nail down good autofocus settings. Looking at my frames, it is clear that sometimes the autofocus procedure does not produce a perfect focus, and the stars are noticeably larger until the next focus. My settings:

Exposure:20 seconds
Full field is the only box checked on the settings page
Annulus from 0 to 100%
Settle time of 2 seconds
SEP detection
kernel size: 3
sigma:1.5
polynomial algorithm
backlash of 100 steps

I am using the ZWO EAF, and switching to SEP and the polynomial algorithm made a large improvement, but there is still definitely room to improve.

I'm also wondering about the twilight settings. I have it set to park the scope once twilight begins, which I love. However, this setting also forces me to delay imaging until well after the sun is down. Maybe at a dark site twilight lasts until 9 PM, but with the amount of light pollution I deal with, it's as dark as it's going to get by 8. Maybe I'm just not seeing the setting, but is there a way to allow imaging to start sooner using the twilight limits?

And then finally, I'm wondering how to warm the CCD after parking. I see there's a setting to do this upon job completion, but I never finish a job in a given night, and the scope parks due to the twilight settings I mentioned above. However, this just pauses the job and the ccd cooler stays on until I manually turn it off in the morning. I want this camera to last for many years, so I'd like to limit the amount of time the cooler is on, especially once summer comes around.

Thanks!

Thanks for reporting this. See if this helps you:

- Change your Ekos mount settings to flip later after meridian.

- Try out the Linear algorithm, and adjust the full-frame ring so that it does not cover the center of the frame, where your object of interest usually stands.

- Twilight has two offsets in the Ekos options, one for dusk and one for dawn. This is not optimal but can help you adjust the beginning and end of your plan.

- From your description you embed your observation in a Scheduler job, thus you should indeed have both parking and camera warm-up at the end of your observation plan, as well as altitude and twilight restrictions for each job (beware of the altitude restriction btw which will reschedule instead of aborting). Note that altitude and parking time in the Mount module are an additional layer of security.

-Eric

Okay, so firstly I hope I didn't accidentally spam this topic. When I posted it I kept getting an error, so I kept reposting until I quit out of frustration. But obviously at least one of those posting attempts went through, so I'm guessing others did too so...sorry about that admins!

In regards to focusing, changing to linear did the trick! I saw this a few minutes ago as I was observing and quickly switched the settings, and sure enough the stars went from flat disks to proper points. I also changed the annulus settings so that the center of the frame and edges can't be used. Now the question is do I throw out all of the data I've gathered these past few weeks because it's been out of focus?...I'm thinking probably not, I'll just fix the stars in post!

I tried having it wait to do the meridian flip for 10 minutes after passing the meridian, but it did not work. However, in the EQMOD settings in the INDI control panel, there is an option that lets you select the target pier side. I switched this to East(pointing west) from auto a few minutes before the flip, and low and behold it flipped with no issues! I see this as a temporary fix though because I'm forcing the mount to be on the east side of the pier, which could be catastrophic if I leave that setting for when I need to point East.

I've located the twilight offsets, thanks! Indeed I do have the observation running through the scheduler, but since I never finish a job in a night, it doesn't warm the camera because the twilight limits just pause it. Is there another way I could run sessions that might circumvent this? Maybe I should just shorten the sequence, but I leave it intentionally long so that the plan doesn't finish while it's still dark.

Quick update: I figured out that I can have it warm the camera and run the observatory shutdown procedure once a sequence pauses if I check the "preemptive shutdown" box in ekos settings. So now all I need to figure out is how to get it to meridian flip on its own.