Calling APT (Astrophotography Tool) gurus - meridian flip failed, need help

[BrendanC]

Hi all,

I recently had a meridian flip fail on me, and I cannot figure out why.

If anyone here is clued up on APT and fancies taking a look at the log and offering suggestions, that would be great - attached.

The fun and games start at the line Starting Automated Meridian Flip at the time stamp 2023/08/17 01:22:27. There are mentions of 'Slew state is false in less than 1s.'. Any recommendations welcome.

Thanks, Brendan

APT_LogFile.log

[AndrewRrrrrr]

you could try the APT forum, I find that the owners are really responsive and helpful.

[BrendanC]

Thanks, I've posted there, not received a reply yet but I agree, Ivo is great so hopefully they'll take a look sometime soon. I'm just trying everything/everyone!

Edited August 18, 2023 by BrendanC

[bottletopburly]

brendan i had same issue  so ivo sent this thread https://aptforum.com/phpbb/viewtopic.php?f=19&t=5101  i didnt have site writes ticked in eqmod ,changed flip moment to 1.0 and delay flip move to 10.0 just read the tips and check your settings ,my last flip worked hopefully next session i can confirm second flip works too. 

[BrendanC]

Thank you.

Edited August 18, 2023 by BrendanC

[teoria_del_big_bang]

A while since I used APT but remember I had a few failures to start with, which was annoying because I thought it was all set up okay as I had stayed up twice to watch the flip previously and all worked flawlessly.

In the end if my memory serves me right it was all down to making sure EQMod was not also controlling the flip which I think is what screwed me but also I think I changed the delays to ensure the flip was always initiated after passing the meridian.
I think if you follow the guide in the link from @bottletopburly you should be fine.

Steve

Edited August 19, 2023 by teoria_del_big_bang

[bottletopburly]

For others reading, the guide in the link is good for whatever image software you use wether it’s Nina , sgp etc as it mainly covers Eqmod settings .

 

Dave 

[BrendanC]

Latest on this is that last night, I increased the Flip Moment from 5 to 10, on a hunch that I may have changed it to 5 a while back and that could be an issue. I tested a flip before the shoot, and it went perfectly - shooting M57, the Ring Nebula. So, I started the main shoot, which was the Dark Shark nebula.

Still, I decided to wait around and make absolutely sure the flip worked - and it failed again. A different fail this time, because I wasn't getting the 1s flip fail errors (which hopefully was because of the increased moment, which stopped tracking etc before hitting the EQMOD limits), but when I went out to check the scope, it was solving on one side of the pier, calculating several thousand pixels out, and slewing across to the other side, flipping each time. I stopped it, waited five minutes, and it was OK. This tells me that I needed to add some time to the Delay Flip Move, but as it was already set to 15 minutes, it's now set at 20 minutes.

It seems to me that 10 minutes Flip Moment plus 20 minutes Delay Flip Move is excessive! That's 30 minutes gone, not including the actual slewing and imaging and solving time.

So, I'm wondering a few things:
* I've been getting large values for the initial offset when plate solving, so my scope must be slightly misaligned compared to where it thinks it's pointing when parked. Would it help to physically try and shift the rig so that I at least get Polaris in my view when it's parked, to reduce this error? Obviously polar aligning too.
* Might the large offset combine with a high altitude object to produce the flip issue? Because the Dark Shark is pretty high in my sky and I can see how small changes can result in large offsets at those altitudes (like the errors you can get if you try to plate solve near Polaris). In which case, could it be that I just need these large moment and delay values for such objects?
* As part of my routine, would it help if I perform a solve sync routine before a shoot, just to help establish a more accurate sky model?

Any/all comments welcome. I'd much rather not have to stay up till 2am again checking that everything is working! @Yoddha, can you shed any insight please?

Thanks, Brendan

[Yoddha]

Hi Brendan,

Could you take a screenshot of the Meridian Flip settings and send me the log file? 20 minutes delay is more than excessive....

Also what are the EQMOD limits? Do you have the right location coordinates set in EQMod? 

The big offset could be caused by misalignment, however the scope is expected to go to the right place as the object is from the same side of the meridian. 

EQMod doesn't like too much sync points located closely, so I would not suggest making solve/sync as part for the main imaging plan. It is critical where the initial sync is made. It must be away from the pole area, declination 85 degrees or less.

[BrendanC]

Hi Ivo,

Thanks for responding.

The EQMOD limits are whatever the default is, I believe just when the RA axis is horizontal.

I'm pretty sure I have the correct location coordinates, as I've been using them for three years with no problem.

I think you're right about misalignment of the mount. This has been brought up elsewhere, so I need to physically move my rig around until I can see Polaris when it's parked, to get the mount closer to where it thinks it's pointing. I'm also getting double diffraction spikes when I stack before and after the meridian flip, which implies cone error. So, I need to sort out my rig!

I'd already sent the log file to your astroplace support email, but I've just sent the very latest (with 2 appended to the file name), complete with screenshot of the meridian flip settings. 

Thanks again, Brendan

[Yoddha]

Hi Brendan,

Many thanks! Looking in the log, there are no signs that you have used Sync. How do you make GoTo alignment?

[BrendanC]

I don't use Sync - I just use GOTO++. That's what I've always done!

If I should be using Sync then that's something I've been doing wrong for about three years...!

[Yoddha]

No wrong thing, just wondering how do you align the scope

[BrendanC]

Phew, good to know!

I think the first thing I need to do, tonight if possible, is to physically align my entire rig so that Polaris is as central as possible when in the parked position. Currently it doesn't even appear in the view when I polar align. This hasn't been a problem before, but clearly it's an issue now, and it's even possible my mount has shifted over time (I keep it outside permanently, under a tarp, and just put the scope on to shoot). Also, as I said, I think I have cone error, which I've never had to address before, so that will be fun and games with Conesharp too.

So many moving parts, so much to go wrong...

[Yoddha]

Clearing the cone error will help to the GoTos and to the flip

[BrendanC]

That's good to know! That's assuming I have cone error, but it does look like that, plus general misalignment of the rig, are at the root of a lot of my problems. I think maybe it just wasn't until I attempted the Dark Shark that these issues came to light.

I still managed to get something though, after combining the good stars with the nebula. Hopefully I'll clear out these other problems in tonight's session.

Thanks for your help again Ivo.

 

[BrendanC]

Actually, very quick question while I'm typing: what would you recommend as starting values for Flip Moment and Delay Flip Move? I can see from the user docs that they're both set to 5 minutes, but Delay Flip Move should perhaps be a bit longer? This is with an NEQ6, 130PDS, ASI533MC Pro, near Oxford in England.

[Yoddha]

You can increase Delay Flip Move to 7 or 10 mins. If there is no danger to hit something you can lower Flip Moment to 2-3 min

As a result, will have almost same waiting time, but will give more time the object to get away from the meridian.

[BrendanC]

So, I had the problem again last night. (calling @Yoddha in case he's around, let me know if the latest log would help cos I'm desperate now...)

The mount kept flipping back and forth, always calculating the object as about 6,000 pixels away.

At one point it stayed on the correct side, but as it solved to get near the object, each plate solve was further away, something like 250 pixels, then 500, then 625 etc.

Finally, after about an hour, after much probing and checking to see what could be wrong, it stayed on the correct side and continued. 

I'm stumped. I ran an alignment plan before the shoot to make sure the mount was properly synced. I could clearly see that the mount was pointing west but it kept flipping. I've checked every setting I can think of, including laptop time settings, EQMOD, APT, I just don't know what's going wrong. I've increased the delay to 20 minutes for next time but I'm pretty sure that's what I had set when it failed previously (I decreased it to 15 hoping the alignment plan would have helped). Any suggestions because this is starting to feel like a hobby killer.

One random thought: is there a case for cone error causing the flip problems? If the scope isn't orthogonal, could this result in enough of a shift in the view after the flip, for the mount to think it needs to flip back to get to the coordinates? I really don't know, it's just a crazy idea I'm putting out here.
 

Edited September 5, 2023 by BrendanC

[Yoddha]

Hi Brendan,

I think that reason for the behavior is the one you suggested - cone error...  It is causing pointing with big difference from both sides of meridian, so big that GoTo++ measures thousands of pixels which the EQMOD takes as position back from the other side of the meridian.

APT is not bug free (as any software more complex than a basic calculator) but the flip is working for many users, which should mean that there is no obvious and general problem...

Maybe this clip could be in help to  fix the cone error - https://www.youtube.com/watch?v=WatdQlPp22Y

[BrendanC]

Thanks Ivo, I'll give it a try. 

[Yoddha]

BTW in the clip Polaris is used for reference, but I suppose that you can use terrestrial objects too... 

[BrendanC]

I'm actually planning on using Robin Glover's Conesharp app.