Auto Center fails after Meridian flip: Page 2

Auto Center fails after Meridian flip

StevenEvan

Well, it has been cloudy here for days and I'm through processing the last clear nights subs, so I have some time on my hands. I'm also interested in this as I've also, just once, had this happen to me. I should have looked into it then but didn't.

So I'll try to make some conclusions from subs 03 and 04, before MF (meridian flip) and after MF.

There are 3 question:

Did the MF actually happen?
Did the AAP plate solve correctly after the MF?
Why is M27 not centered?

1. Did the MF actually happen?
I ran both sub 03 and 04 through a online plate solver and the appended images are:

Sub 03

Sub 04

The star HD 189657 location does change in reference to the M27 and the image frame.
So Yes, the MF took place

2. Did the AAP plate solve correctly after the MF?
This is interesting. Sub 04 plate solve shows the image center as:
RA: 19h 59m 32.945s Dec:+22° 22' 13.348"
The actual location of M27 (in AAP database) is:
RA: 20h 0m 6s DEC+22° 46' 53"

Given that the pixel scale is 0.47 arcsec/pixel this means that M27 should be roughly 3000px left or right of center and around 70 pixels up/down. Which is what we see in the image.

The interesting thing is that this is where AAP believed the center to be.
In your log is shows that after MF the AAP believes the target is at:
RA: 20h0m22s DEC:+22°25'54"
and tried to plate solve to this RA/DEC and it succeeded within around 400 pixels.

In short, the plate solves after MF ended up at the location that AAP was looking for.
So also Yes, AAP plate solved to this location correctly. It got to the location where it thought the image was centered.

3. Why is M27 not centered?
So what we have here is AAP not knowing the correct image center. During MF it used an image center that was not correct. It didn't go the the wrong RA/DEC, it just didn't know that you had M27 in the image center.

Assuming that this is not a software bug in AAP then I can think of one way this may happen.
I GOTO M27, its not centered so I use plate solve and SYNC to update the AAP RA/DEC.
I then move the mount some way that AAP doesn't know about, maybe the mount hand controller to center M27 in the image. (Note that AAP is not aware that the image center has changed.)
I begin AutoRun.
MF happens and AAP GOTO and plate solve is going to go to the original RA/DEC, not the centered image location.

Does that all make sense?

Steve

WalterT

w7ay 1 degree in the east-west direction will cause the mount to think the target crosses the Meridian either 4 minutes

StevenEvan I GOTO M27, its not centered so I use plate solve and SYNC to update the AAP RA/DEC.
I then move the mount some way that AAP doesn't know about, maybe the mount hand controller to center M27 in the image. (Note that AAP is not aware that the image center has changed.)
I begin AutoRun.

Guys, thank you very much for all of the great tips. A lot of the things you guys analyzed are way over my head but based on your feedback I think I narrowed it down to either a leveling problem or a problem of me re-centering the image manually after the ASIAIR go-to, or both. So when getting ready for imaging I create my Autorun plan, search for my object in the AAP, and hit the go-to button. The AAP will slew to the target, plate-solve 1 or 2 times and then settle when it thinks the target is centered. Most of the time the target is not centered so I manually move my target to the center as Steven said, then start guiding, then start my Autorun. So obviously, as Steven mentioned, if my Autorun includes a MF then it will go to were it had originally thought the center was and not to where I moved it afterwards.
Also, I pretty much took it as a given that my mount was not perfectly leveled, but I assumed that a little error wouldn't be a big deal since I have plate solving and guiding. But, going off what Chen said that just a degree off on leveling can cause up to 4 minutes in discrepancy I can see how that could cause big problems.
So last night I put the mount on bricks and off the grass, it's all I have on hand right now, and did my best to level it with the cheapo bubble level built into the mount. I also went to my mount Meridian settings and changed the times to stop tracking and flip after from 1 minute to 5 minutes, just to give objects more time to pass the meridian to compensate for errors in my leveling. I tested the system by running a test run on Vega. When the mount settled on Vega it was not centered but it was not too bad I would say about 1/8 of the width of the sensor away from the cross-hairs. So I just left it and ran it and it worked, it flipped after 10 minutes and it settled on the other side in the same slightly off center Vega. I did the same thing for my Autorun of M27, and this was one of the rare times when the AAP actually put the object in the exact center of the frame so I was not tempted to make any manual adjustments. I run the sequence and woke up to a successful meridian flip. So in short, either my leveling error or my re-centering the frame manually after the AAP auto-center or my MF timing settings were causing the problem. I'll just need to incorporate these things into my routine from now on, which is a lot less hassle than waking up to half a night of ruined subs.

Here are the low-res thumbnails of image from before and after the flip.

ps. You got it Chen, RTV silicone. We sell a lot of Permatex products at work so I get a good discount on the stuff :

wvreeven

If you use the hand controller to center to a certain position then that very well may explain why the centering is "off" after the meridian flip. AAP is leading in the position that the mount points at.
I wasn't aware of this until I read another comment of Chen to another thread, but you can tap (or click) at a position in an image taken with AAP and then click GoTo to center a certain position. I'd advice you to do that next time you want to center an object because that way AAP at least knows what position the mount should be pointing at.

w7ay

wvreeven next time you want to center an object because that way AAP at least knows what position the mount should be pointing at.

Personally, I do not use that feature myself :-). I use the coordinates from SkySafari (one of the few instances I find SkySafari useful; I can do without the eye candy otherwise), for example, find a coordinate between M81 and M82, and enter that into the mini-GOTO window in ASIAIR -- which you can also save as a Favorites if you want to go to the same location in the future; I have a BodesNebulas entry :-).

Heck, I don't even use the ASIAIR Auto-Center function. I do my own rinse-and-repeat plate-solve-Sync-And-GOTO-capture-new-image, since ASIAIR's auto-center is very coarse; perhaps only good enough for people who don't know any better.

Chen

wvreeven

Yes that's a possibility as well. I use SkySafari as well and that works really well. SkySafari sends position commands to AAP which then commands the mount which is why this works very well. My point was that using the hand controller probably doesn't work if the AAP is leading in "knowing" the position that the mount is pointed at.

w7ay

WalterT I filled mine with silicone glue after I screwed on the boot so sounds like I was on the right track.

By the way Walter (at least I have no fear of hijacking this thread, since it is yours :-), if you mean the Silicone RTV, that is fantastic material. Another material that resulted from the NASA space initiatives, I think. And yes, they are electrically inert after curing.

I use RTV in a number of places too, for both of my hobbies. (My newfound favorite "glue" is the Bondic UV curing stuff :-).

Back in grad school the early 1970s, I was involved with an (US) NSF project for remote satellite based weather sensors for deployment to the Antarctica (they transmitted data autonomously to Nimbus and Tiros low-earth orbit satellites that dumped the data when they passed Alaska).

We used the RTV to hold down wiring, and as far as I know, it held up for decades. I had seen temperatures below -40ºC transmitted back when we were monitoring it, before handling the system to hardcore weather researchers; Antarctica used to be a void of data for global weather modeling. I was with the school's Radar Astronomy group -- we did radio, not weather :-).

Chen

w7ay

WalterT a leveling problem or a problem of me re-centering the image manually after the ASIAIR go-to ... Most of the time the target is not centered so I manually move my target to the center as Steven said, then start guiding, then start my Autorun

Don't do that with your hand controller, Walter. And if you do it from the ASIAIR directional buttons, you would need to sync the information back to the mount.

If you want to see what people go through leveling their mounts, take a look at this YouTube video (I have the same mount).

https://www.youtube.com/watch?v=7IC1j8i-1H4

I use a two axis inclinometer (with 0.01º precision, but don't know what its accuracy is), so I don't have to keep moving the inclinometer's orientation around :-).

Good luck.
Chen

StevenEvan

@WalterT you're right, there are definitely 2 "takeaways" from this thread.

Firstly, setup the mount as level and stable as possible. You have to "build on a good foundation".

Secondly, make sure that AAP knows where the mount is pointing.

Use AAPs direction buttons
Or hold your finger on the image location you want to center until the green target appears
Or if you really have to use your hand controller make sure you plate solve and sync after moving.

Then AAP will know the image center.

Steve

wvreeven

I'd like to add (or repeat even, I have written this before) that I never level my mount and I never have centering issues. Plate solving works with RA and DEC and it will continue moving the mount until the target RA and DEC are close enough to the center of FOV. This is unaffected by how level the mount is provided the mount has been polar aligned well enough.
The reason why the mount was "off" after meridian flip was the use of the hand controller. Just only use the AAP for framing (either via the app or via SkySafari) and you'll be fine.

WalterT

wvreeven Thanks wvreeven. I'm going to have to check out SkySafari, from everyone's comments it sounds like it is a lot more convenient. I had already purchased Stellarium but that was before I knew SkySafari could connect directly to the ASIAIR.

Rocaillou

I would like to re-open this discussion here for simple reasons :

Leveling is important but not the point mentioned here. My leveling on EQ8 is good.
ASIAR does the job at any time (solving + centering) and perfectly

But in the process of meridian flip solving+centering, the target is off by a few degrees.

I have increased to 15 minutes the after meridian flip to start but no difference.
Taking the images un-centered and asking a platesolve+ centering and of course the target ends up right in the middle where it should be.

To me, it was working absolutely fine, until it stopped working fine for a few weeks now, and all my “after meridian flip”images are to be trashed. i am loosing energy in finding the solution.

I use ASIAIR connected directly to the mount (EQ8), a 12”SCT with ASIAIR OAG with the ASI6200 MC @F10.
Adding below the log of ASIAIR saying meridian flip work and solving/center all good.

I dont see here any comments from ZWO as i am convinced something changed in latest released which was working fine.

Can someone that had this problem and manage to solve it help me please or best to have ZWO answer ?

Thank you

wvreeven

w7ay Now wait for the star to transit just past the Meridian. This happens at exactly when the Local Sidereal time is equal to the Right Ascension of the star. At this point in time, issue another GOTO.

There are several mount brands (like SkyWatcher and iOptron) that only perform a meridian flip several minutes after culmination. You need to take this into account since this is an EQ8 which, unless I am mistaken, is a SkyWatcher mount.

Rocaillou I use ASIAIR connected directly to the mount (EQ8)

How do you connect the mount exactly? Using a cable between the ASIAir and the hand controller or via an EQMod cable or yet via another way?

Rocaillou

No reply to my additional post… please help> see below my bubble level which is fine…!

w7ay

Rocaillou

Check your mount without ASIAIR.

Use its hand controller to GOTO a star (or coordinate) that is just to the east of the Meridian for a star on the hemisphere the Zenith is at. A star is easier to visualize, but any coordinate would do.

Use the GOTO command to move to that coordinate (do not slew -- use the GOTO -- since that should also turn tracking on with pretty much every mount).

Now wait for the star to transit just past the Meridian. This happens at exactly when the Local Sidereal time is equal to the Right Ascension of the star. At this point in time, issue another GOTO.

The second GOTO should cause your mount to perform Meridian Flip. There are a few mounts (e.g., the old Avalon StarGO) that does not have a Meridian Flip function; if so ASIAIR will not Meridian flip either -- ASIAIR uses the Meridian Flip in the mount.

If this second GOTO does not perform a Meridian Flip, you need to fix the problem. It is completely independent of the ASIAIR.

If the above works, then you need to make sure the ASIAIR and the mount maintains the correct times. The Local Sidereal Times must agree. Since you cannot check LST in ASIAIR, you can equivalently check UTC and Longitude. If your mount has GPS, be sure to turn it off and let the ASIAIR push the UTC and Longitude values from the tablet to the mount.

By the way, the level in mounts are very rough. Use a long bubble level, or a digital inclinometer. Tell the mount to move the hour angle to zero, and turn off tracking, and set the declination to your latitude -- i.e., the OTA should now point to Zenith. Now, use the bubble level along the side of the OTA to check if the OTA is perfectly perpendicular.

Chen

Tom62e

w7ay
Sir - if this is all true (about the levelling being the culprit) than why does it center properly in the beginning of the session when I GoTo my target.

I should note that my case is slightly different than the OP's because his initial Goto was not accurate and he was manually making the correction to center it properly.

That is not the case for me. My initial GoTo command centers fine. It's only after the Meridian Flip that it won't center properly.

Do you still think this could be a levelling issue? I'll double check the levelness of my pier but it's always been perfect so I doubt that has changed.

Thanks,
Tommy

w7ay

Tom62e I should note that my case is slightly different than the OP's because his initial Goto was not accurate and he was manually making the correction to center it properly.

Your situation is one where the OTA is not pointing at Zenith when the pier side changes. I suspect your meridian flip is happening at the correct time, unlike the OP's problem.

In your case, try a meridian flip at the Celestial Equator vs a meridian flip closer to the pole. The different behavior might also give you a clue.

You may have a "cone error" (i.e., when pointed east or west (0 declination), the OTA is not perfectly orthogonal to the RA axis). Things like that can be caused by the dovetail clamps/plates on the declination axis, or when the OTA not perfectly parallel to its dovetail plate.

Chen

« Previous Page