Hi all,

Last night I was imaging M83 and my mount did a meridian flip just before midnight. The flip went off without a hitch except for the guiding. Basically when it captured and sub-framed it didn’t pick up a star. I had to stop the calibration and manually pick a star for it to guide on.

Are there any settings I should look at so that it works next time?

Regards

Rene

Hi Rene,
the easiest way for us helping you are log files. Do you have them at hand? If yes, please upload them.

If not, please describe in more detail. E.g. are you using the internal guider or PHD2?

- Wolfgang

So I managed to repoeat the issue last night while imaging Cent A. Meridian flip completed, post flip alignment was fine, but it failed on auto-guiding. Looks like it couldn't find a guide star and just gave up. I was using the internal guider. The relevant part of the log file is below and I've attached the full file. There's also a screenshot of the guiding module which is complaining about mount back-lash. Never seen this error before so figure it's just a generic "it failed and I can't be bothered recovering gracefully" error.

[2020-04-22T23:29:00.787 AUS Eastern Standard Time DEBG ][ org.kde.kstars.ekos.guide] - Iteration # 21 : STAR 131.712 , 90.1625
[2020-04-22T23:29:00.793 AUS Eastern Standard Time DEBG ][ org.kde.kstars.ekos.guide] - Iteration 21 Direction: RA_DEC_DIR Duration: 2000 ms.
[2020-04-22T23:29:00.799 AUS Eastern Standard Time DEBG ][ org.kde.kstars.ekos.guide] - Star x pos is 29.7237 from original point.
[2020-04-22T23:29:00.804 AUS Eastern Standard Time DEBG ][ org.kde.kstars.ekos.capture] - Guiding state changed from "Calibrating" to "Calibration error"
[2020-04-22T23:29:00.810 AUS Eastern Standard Time INFO ][ org.kde.kstars.ekos.capture] - "Post meridian flip calibration error. Aborting..."
[2020-04-22T23:29:00.816 AUS Eastern Standard Time INFO ][ org.kde.kstars.ekos.capture] - "CCD capture aborted"
[2020-04-22T23:29:00.823 AUS Eastern Standard Time DEBG ][ org.kde.kstars.ekos.capture] - setMeridianFlipStage: 2
[2020-04-22T23:29:00.829 AUS Eastern Standard Time INFO ][ org.kde.kstars.ekos.guide] - "GUIDE_RA: Scope cannot reach the start point after 21 iterations. Possible mount or backlash problems..."
[2020-04-22T23:29:00.837 AUS Eastern Standard Time DEBG ][ org.kde.kstars.indi] - ZWO CCD ASI120MM Mini : "[DEBUG] Uploading file. Ext: fits, Size: 72000, sendImage? Yes, saveImage? No "
[2020-04-22T23:29:00.842 AUS Eastern Standard Time DEBG ][ org.kde.kstars.indi] - ZWO CCD ASI120MM Mini : "[DEBUG] BLOB transfer took 0.00132023 seconds "
[2020-04-22T23:29:00.848 AUS Eastern Standard Time DEBG ][ org.kde.kstars.indi] - ZWO CCD ASI120MM Mini : "[DEBUG] Upload complete "
[2020-04-22T23:29:01.252 AUS Eastern Standard Time DEBG ][ org.kde.kstars.indi] - ZWO CCD ASI294MC Pro : "[DEBUG] Aborting camera exposure...

This happens to me almost all the time. The only time the SUBFRAME and AUTO STAR options together works, is by pure random luck. I can't see any cause and effect as to why it fails.
And not just with automated flips, but anytime I try to use these features. EKOS just seems to pick a random spot, usually with no stars in it, and tries to guide on it., This is with UBUNTU and its my MacOSX versions.
For the life of me, I have no idea how others are getting it to work. It has not functioned for me for almost two years, and I keep my software to to date.

When I do use the automated meridian flip, I have better luck if I turn off the "Subframe" option. Then I get about a 75% chance of finding a guide star. Otherwise it's about 10% with the subframe on.

Peter

For me it works by enabling "Reset Guide Calibration After Each Mount Slew" by clicking on Options to the lower right of the guide tab:




Wouter

Hi Peter,

I saw your Facebook post and was the one who asked the questions. Some good tips there. I'll try tonight.

From what you and others have said, the guide module isn't fool-proof after a meridian flip. Perhaps the devs need to look at it for the next release. It needs to be more robust when it fails to find a guide star. Not just give up.

Rene

I'm hoping after the sextractor changes by Robert makes it into KStars, we'd have much better luck in detecting stars than before.

So the meridian flip failed again last night. This time during alignment. From what I could see the solver failed and just gave up. Surely it should be able to reset itself and try again. There’s far too much manual intervention and monitoring required.

I’ll pull the logs and attach them for anyone who is interested.

Rene

Ok here's the log file (attached)

I've had a quick read of it and you should start looking from the following timestamp:
[2020-04-25T02:28:56.092 AUS Eastern Standard Time INFO ][ org.kde.kstars.ekos.capture] - "Telescope completed the meridian flip."

From this point the alignment module kicks in and tries to solve. It attempts to re-align for the rest of the night with no luck. The last "Solver Failed" mesage I can find is at timestamp:

[2020-04-25T02:41:41.729 AUS Eastern Standard Time INFO ][ org.kde.kstars.indi] - Astrometry : "[INFO] Solver failed. "

Then the mount/Ekos sits in a never ending loop of aligning until I killed it at 5:00am this morning.

Cheers

Rene

This is what is strange, because last night mine was successful again.
Ever since I made sure the "solve and slew" button was set, and that I turned off "Subframe" in the guider, my success rate went way up.

Peter

I never take mine off "solve and slew" and I turned off sub-framing. Perhaps it's because I'm running Kstars on a Windows machine instead of from the RPI.

I'm becoming so frustrated with the constant issues. I know the software is free but really it should be able to handle something as simple as move the telescope, take a picture and compare it to a file, then rinse and repeat until successful. Might be time to consider another software package like SGP or Voyager.

Rene

I’m in a similar frustration boat, in that every new update breaks something else that worked before.

But I hate using MS Windows and while I put up with it at work, won’t have it in my house. The Mac OSX version works better than the Linux version so once I finish my observatory I will bypass the Linux machine I use as a bridge and direct connect a Mac to my equipment. My new mount (CEM120 EC2) will make that easy with its built in USB ports right at the dovetail plate.