Guiding failed after Meridian Flip

[CraigD1986]

I was shooting M101 last night and my ASIAir and EQ6-R Pro did a meridian flip at about 2am. After that point, all of my images were unusable with massive star trailing. Here is what the log shows:

2023/04/04 02:11:35 [Meridian Flip|Begin] Wait 10min43s to Meridian Flip
2023/04/04 02:22:18 Meridian Flip 1# Start
2023/04/04 02:22:18 [AutoCenter|Begin] Auto-Center 1#
2023/04/04 02:22:18 Mount slews to target position: RA:14h4m2s DEC:+54°14'18"
2023/04/04 02:23:12 Exposure 2.0s
2023/04/04 02:23:15 Plate Solve
2023/04/04 02:23:16 Solve succeeded: RA:1h57m15s DEC:+85°38'13" Angle = 61.8695, Star number = 208
2023/04/04 02:23:16 [AutoCenter|End] Close to Celestial Poles, unable to Sync
2023/04/04 02:23:16 [Meridian Flip|End] Meridian Flip succeeded
2023/04/04 02:23:16 Start Tracking
2023/04/04 02:23:16 Wait for Mount Settle
2023/04/04 02:23:26 [Guide] ReSelect Guide star
2023/04/04 02:23:27 [Guide] Start Guiding
2023/04/04 02:23:31 [Guide] Guide Settle
2023/04/04 02:23:36 [Guide] Settle Done
2023/04/04 02:23:36 Exposure 120.0s image 148#
2023/04/04 02:25:38 Exposure 120.0s image 149#
2023/04/04 02:27:39 [Guide] Stop Guiding
2023/04/04 02:27:39 [AutoCenter|Begin] Auto-Center 1#
2023/04/04 02:27:39 Mount slews to target position: RA:14h4m2s DEC:+54°14'18"
2023/04/04 02:27:44 Exposure 2.0s
2023/04/04 02:27:46 Plate Solve
2023/04/04 02:27:48 Solve succeeded: RA:1h57m14s DEC:+85°37'58" Angle = 61.882, Star number = 202

2023/04/04 02:27:48 [AutoCenter|End] Close to Celestial Poles, unable to Sync
2023/04/04 02:27:52 [Guide] ReSelect Guide star
2023/04/04 02:27:53 [Guide] Start Guiding
2023/04/04 02:27:57 [Guide] Guide Settle
2023/04/04 02:28:08 [Guide] Settle Done
2023/04/04 02:28:08 Exposure 120.0s image 150#
2023/04/04 02:30:10 [Guide] Dither
2023/04/04 02:30:10 [Guide] Dither Settle
2023/04/04 02:31:10 [Guide] Settle Timeout
2023/04/04 02:31:10 [Guide] Stop Guiding
2023/04/04 02:31:12 [AutoCenter|Begin] Auto-Center 1#
2023/04/04 02:31:12 Mount slews to target position: RA:14h4m2s DEC:+54°14'18"
2023/04/04 02:31:17 Exposure 2.0s
2023/04/04 02:31:20 Plate Solve
2023/04/04 02:31:21 Solve succeeded: RA:1h57m16s DEC:+85°37'53" Angle = 61.8854, Star number = 202

Then this just repeats for the rest of the session. I'm not sure what happened or how to avoid this next time. I think I had the option to re-calibrate guiding after a flip turned off as this caused problems last time.

[The Admiral]

When did you last do a meridian flip, and did it work then?

Ian

 

[CraigD1986]

The night before and it worked as normal

[ONIKKINEN]

Looks like the mount never actully slewed all the way during the meridian flip. DEC 85 and RA 1h57m is on the wrong side of Polaris as the intended target, but why exactly the mount wouldnt slew again is beyond me. Looks like some kind of setting that prevents syncing close to the poles (but then again the mount is not suppposed to be there).

Could the scope hit the tripod when pointed towards the zenith? That way the motors would start skipping steps leading the mount to think it has moved the scope when in reality it was stuck against the tripod for half of the slew, which could place the scope close to the pole where the "unable to sync" thing comes out.

[symmetal]

It seems the mount, for some reason, stopped moving in RA and Dec at about the same time half way through completing the flip. As it ended up near the pole the sync was not performed after the plate solve as it's known the result would not be accurate. The ASIAir seemed to assume it an error in syncing, and ignored it, and declared this as a flip completed, so started imaging again. Every few minutes later, presumably before each new exposure, the mount  is commanded to slew to the new target position again as the ASIAir presumably hasn't had confirmation that it is on target. The mount doesn't actually move at all each time so the same response is given and the process just repeats. 

It looks like the mount shut off power to the motors mid way through the flip due to some failure detected, possibly power related, and never restored motor power. A sudden increase in current drawn, due to the scope being prevented from moving, by hittting the tripod or a cable snag may cause the mount to enter a fail safe mode where a repower is needed to reset it.

Alan

[Rallemikken]

On 04/04/2023 at 16:51, CraigD1986 said:

Then this just repeats for the rest of the session. I'm not sure what happened or how to avoid this next time.

Maybe there will not be a next time. This has happened to me once, during 100+ sessions the last three years. Never found out what was happening, and it scared me, but it never repeated. I run Ekos/INDI, belive that your rig does the same.

I was nearby and just turned everything off, and rebooted the system. All OK.

Edited April 5, 2023 by Rallemikken

[CraigD1986]

Thanks all for the input here.

I was using a Skywatcher 190MN, which is a big scope but I’ve just checked my CCTV camera footage and it doesn’t look like there was a collision during the flip (or at any other point).

In terms of power, the mount and ASIAir are running on mains via a regulated DC power supply (Nevada PS-08) so I wouldn’t have thought it could be a power issue.

Cone to think of it, when I was setting up, I polar aligned, slewed to M101 and then released the RA clutch, rotated the mount to check for collisions with the back of the scope. Then when I went to slew to M101 again via SkyView on ASIAir, everything was way off, including when I told the mount to go home. I then manually pointed at Polaris, plate solved and pressed “sync to mount”. Then I slewed to M101 successfully and began imaging. I wonder if something was out at this point, causing the incorrect position after the flip?

[CraigD1986]

Also, the ASIAir was setup to return the mount to home at the end of the plan. In the morning, the RA was in the home position but not the declination. 

[symmetal]

19 hours ago, CraigD1986 said:

Cone to think of it, when I was setting up, I polar aligned, slewed to M101 and then released the RA clutch, rotated the mount to check for collisions with the back of the scope. Then when I went to slew to M101 again via SkyView on ASIAir, everything was way off, including when I told the mount to go home. I then manually pointed at Polaris, plate solved and pressed “sync to mount”. Then I slewed to M101 successfully and began imaging. I wonder if something was out at this point, causing the incorrect position after the flip?

Releasing the clutch and rotating the mount manually will make it lose knowledge of where it's pointing. As far as the mount's concerned it's still pointing at M101. The AZ-EQ6 has encoders on each axis which will tell it if it's been moved manually, but the EQ6-R doesn't. When you went to re-slew to M101 it likely didn't move much as it thought it was already pointing at it. Hence it would be way off when you homed it, as it thought it was homing from being pointing at M101.

Using Polaris to plate solve is not recommended for an eq mount, as unless your polar alignment is spot on, as it will likely give a RA position thet is significantly off being correct, because the RA hour lines are very close together near the pole. If you're pointing at the pole itself, any random RA position will be correct.

I don't know if the ASIair keeps an internal sky map of its sync points to assist with slewing if your polar alignment is not perfect, similar to what a two or three star alignment does. If it has sync points for M101 and an incorrect Polaris it would make the position after the flip, off to some amount, but unlikely to be as far off as you had. It's only because you unfortunately ended up near the pole it wouldn't accept the plate solve as being accurate, so wouldn't then continue slewing to the correct position of M101.

EQMod has an annoying feature where it will maintain the previous sync points between sessions, and you have to manually delete them at the start of each session, or your slews will likely be off and it can sometimes never centre to the correct position.  If plate solving is used, keeping a sync point map is not a benefit. I assume ASIair doesn't keep sync points between sessions so this shouldn't be an issue for you. 🙂

Alan

[CraigD1986]

I think we may have borrowed down the problem here then. It looks like this was caused by user error. So in future, if I release the clutch and manually move the mount, should I just plate solve at that point rather than returning to home first?

That said, the mount did find M101 successfully when I started the imaging session as I got over 4 hours of good data before the flip. It was only after the flip that this was a problem.

[symmetal]

If you had previously plate solved before manually moving the mount then plate solving again will end up with two sync points that don't correspond with a valid map of the sky as the mount is unaware of how much it was moved manually. The mount just computes the number of stepper motor pulses to apply to go from one location to another. This will likely affect the home command too so it will not return to the correct home position. Doing a Park and Unpark should remove any sync points I would expect, or else it wouldn't fix the problem.

If you haven't plate solved before manually moving the mount, then plate solving after moving should be OK. The mount knows the RA and Dec coords at the home position, so can work out how many steps are needed to get there from the plate solve. You can easily check this by doing a home command after this first plate solve after manually moving, and check it does in fact go home correctly.

It found M101 correctly for your first session, but how far off was the initial slew before it had to do more plate solves and slews to get correctly on target. If your flip hadn't ended up near the pole, then it may have completed the flip correctly with a few more target centreing slews and you wouldn't have known there was an issue. I don't know how far ASIAir will allow a first slew to be off target before it aborts any further slews. I use SGP with ASTAP plate solving, and SGP can report a failure and halt if the first slew is too far off target.

Alan

Edited April 6, 2023 by symmetal

[CraigD1986]

I thought the "sync mount" option was supposed to clear all of these settings and reset everything?

[symmetal]

7 hours ago, CraigD1986 said:

I thought the "sync mount" option was supposed to clear all of these settings and reset everything?

If the ASIAir allows you to do a two or three star alignment then it means it has the capability of keeping its own internal sync point map of the night sky to correct for polar alignment errors. There may be an option to sync the mount with every successful plate solve which you can perhaps turn off, or to have an option not to maintain an internal map of successive sync points and just use the last one, which is the better option if you're always using plate solving. Another ASIAir user may be able to say what options are available.

Alan

[jhayton]

I recently had this issue pop up as well.  Ive been imaging NGC 6888 for several days in a row since June 27th.  The first 3 days the auto run sequence went off without a hitch.  Nights 4,5 and 6 have had issues after the Meridian flip.  Images are streaked as if no tracking.  

Looking at the log, the only notable difference I see is guide did not settle and timed out.

I'm on a semi permanent setup.  Nothing changed with the rig during this time.  Only thing noteworthy is I used a new Android device on day 3.  I normally run from my iPad.  Days 4 onward I was back to my iPad, and thats when I see issues.  I feel something got changed in the guiding settings, but it all looks right.

Sky-Watcher CQ350
ASAair MINI

(abnormal session on the left, normal on the right)

[Elp]

There are a few threads on the ZWO forum regarding failed meridian flip and plan continuation. The last one I read mentioned that all dates and times and UTC time offsets to be correct across devices (hand controller, phone) which the asiair communicates with even to the point of disabling DST on the hand controller. It apparently doesn't like daylight savings time being applied. This was more to do with the flip not starting though.

The issue here I suspect is something else, I've had random slewing, guiding issues, home position resetting mid plate solve altering the sky position alignment and tracking turning itself off since firmware 10.47 or so. Next session it goes away without changing a thing, then something else goes wrong.

I've persevered with it, but only by using an older controlling app and firmware version on the air.

 

 

Edited July 7, 2023 by Elp

[CraigD1986]

On 06/07/2023 at 20:24, jhayton said:

I recently had this issue pop up as well.  Ive been imaging NGC 6888 for several days in a row since June 27th.  The first 3 days the auto run sequence went off without a hitch.  Nights 4,5 and 6 have had issues after the Meridian flip.  Images are streaked as if no tracking.  

Looking at the log, the only notable difference I see is guide did not settle and timed out.

I'm on a semi permanent setup.  Nothing changed with the rig during this time.  Only thing noteworthy is I used a new Android device on day 3.  I normally run from my iPad.  Days 4 onward I was back to my iPad, and thats when I see issues.  I feel something got changed in the guiding settings, but it all looks right.

Sky-Watcher CQ350
ASAair MINI

(abnormal session on the left, normal on the right)

The first time I had an issue like this, i was told to not re-calibrate the guiding after a flip so I turn this off in the settings and then everything worked as expected again.