×

INDI Library v2.0.6 is Released (02 Feb 2024)

Bi-monthly release with minor bug fixes and improvements

Meridian flip

  • Posts: 245
  • Thank you received: 56

Replied by Rishi Garrod on topic Meridian flip

I am also looking forward to the answer to this from someone how knows how this actually works.

My own observations are that if you just have a Sequence running then the mount will flip but imaging will not continue correctly. You really need to do a new plate solve and restart the tracking. If you use the Scheduler then it takes care of plate solving, re-focus if required and tracking.
2 years 4 months ago #77781

Please Log in or Create an account to join the conversation.

  • Posts: 1208
  • Thank you received: 559

Replied by Hy Murveit on topic Meridian flip

@Morelli,

Thanks so much for documenting the system! We really need good doc of Ekos for all languages. 

Wolfgang is the expert on this topic, meridian flips, but I'll take a shot at my understanding of what's going on to get you started. (I'll try my best--but please follow up if I don't respond to what you're looking for.)

Ekos, inside the mount module, regularly monitors the telescope position that your mount hardware broadcasts. If enabled, the mount module will try to initiate a meridian flip once the hour angle condition for a meridian flip is met. This condition is something like "mount on west side of pier, pointing east, and hour angle is positive by a specified amount, e.g. the parameters filled in on the mount tab".  The mount module communicates with the capture module, so if, when the condition is met, and a capture is in-progress, it will wait until the end of that capture.

When the mount module is ready for a meridian flip, and capture is no longer busy, the capture module suspends guiding and tells the mount module to go ahead with the flip.

The slewing part of the meridian flip is really performed by the mount. That is, all Ekos (the mount module) does is issue a slew command to the target position. The target position is usually the current position of the mount give or take a little bit of dithering, guide error, and alignment error. It is basically the desired position from the previous slew. So, all Ekos does is issue a slew command to "the target position (RA/DEC)". It is the mount hardware, not Ekos, which should realize "the target position should really be pointed to from the other side of the pier", and should slew over there.  Most meridian flip issues wind up being caused by this (e.g. the mount not realizing that the pier side should be changed, e.g. because it doesn't think it's near hour-angle=0, or because it has the wrong lat/lon, wrong time, etc). This is why we have meridian flip conditions of a certain margin past the meridian, to insure that a small difference between the mount's position understanding, and ekos' doesn't cause the meridian flip to fail.

When the meridian flip slew is done, and the mount seems to have changed pier sides, the capture module is in charge of resuming processing. It resumes guiding, if guiding was suspended -- i.e. if guiding wasn't running before the flip, it won't start guiding after the flip. It starts up an alignment for the target position. I don't believe there are any parameters that need to be set for it to resume guiding or start alignment. Once those things perform successfully, capture resumes its capture sequence.

Hy
2 years 4 months ago #77783

Please Log in or Create an account to join the conversation.

  • Posts: 437
  • Thank you received: 31

Replied by Paul on topic Meridian flip

Morelli,

You may want to refer to the discussion I have had recently on Meridian flips and the limitations with the current implementation.

indilib.org/forum/mounts/10641-meridian-....html?start=12#77673

Paul
Last edit: 2 years 4 months ago by Paul.
2 years 4 months ago #77792

Please Log in or Create an account to join the conversation.

  • Posts: 335
  • Thank you received: 19

Replied by MORELLI on topic Meridian flip

@Paul: I read your post before.
2 years 4 months ago #77841

Please Log in or Create an account to join the conversation.

  • Posts: 300
  • Thank you received: 57

Replied by Scott Denning on topic Meridian flip

For me, the key to successful meridian flips has always been to make the mount settings (either in the driver or the hand control) consistent with the way I'm asking for a flip in Ekos.

For example, set the meridian limit in the mount to something small, like 1 degree past the Meridian. This tells the mount to flip if a GOTO is executed to coordinates more than 1 degree past the meridian

Then in the Ekos scheduler, set the Meridian Flip to take place 3 degrees past the Meridian.

The mount tracks past the meridian, taking exposures. When it reaches 3 degrees west of the Meridian, Ekos issues a GOTO to the RA/DEC to which the mount is already pointing. As Hy says, it's the MOUNT (not Ekos) which then executes the flip.

This example shows why the Ekos flip setting must be FARTHER WEST than the mount Meridian Limit. The important thing is to coordinate the TWO settings: one in the mount itself and the other in the Ekos Scheduler.
Last edit: 2 years 4 months ago by Scott Denning.
2 years 4 months ago #77843

Please Log in or Create an account to join the conversation.

  • Posts: 992
  • Thank you received: 155

Replied by Peter Kennett on topic Meridian flip

You state: "...set the meridian limit in the mount to something small, like 1 degree past the Meridian. This tells the mount to flip if a GOTO is executed to coordinates more than 1 degree past the meridian. Then in the Ekos scheduler, set the Meridian Flip to take place 3 degrees past the Meridian."

In my case, with the CEM120 EC2, the exact opposite is needed. If I set my mount to flip 1 degree past Meridian with the mount's hand controller, it will flip 1 degree past meridian, regardless of what EKOS is doing. If I tell EKOS to flip 3 degrees after, it will never get executed, because the mount already flipped at one degree past. This means my exposures are messed up at 1 degree past, since EKOS thinks it can wait until 3 degrees past.

I have to set my mount further west than EKOS, and it becomes a failsafe in the event EKOS doesn't flip for some reason.
Last edit: 2 years 4 months ago by Peter Kennett.
2 years 4 months ago #77850

Please Log in or Create an account to join the conversation.

  • Posts: 437
  • Thank you received: 31

Replied by Paul on topic Meridian flip

Scott,

I think you are overly simplifying thiings.
You need to set the mount up and then configure the software to meet the requirements of the mount.
The problem is that a mount has a range over which a goto will produce a meridian flip but the software is not capable of recognising a range only a fixed limit.
So, you need to set the software to produce a flip after the first point at which it will flip but before it reaches the limit.
Since, Ekos only has one setting it needs to be set to ensure it always commands the goto before it reaches the limit.
For long exposures it may happen that it cannot achieve both and will sometimes fail, depending on when a previous exposure completes.
I have had some discussions with the programmers and they are considering adding an extra setting to fix this but it has not currently been implemented.
So in summary, to ensure 100% flips working you need to have the exposure length shorter than the period between the first point and the limit and to ensure the goto is commanded before the limit is reached.

Paul
 
2 years 4 months ago #77854

Please Log in or Create an account to join the conversation.

  • Posts: 992
  • Thank you received: 155

Replied by Peter Kennett on topic Meridian flip

That is especially important with long exposures - like over 5 minutes long. Your idea is what NINA does. You set the range, and if an exposure would go over the limit, but it is too early to flip, it will pause the exposure sequence until the mount is capable of flipping. This prevents the mount from forcing a flip during an exposure.
2 years 4 months ago #77856

Please Log in or Create an account to join the conversation.

  • Posts: 300
  • Thank you received: 57

Replied by Scott Denning on topic Meridian flip

Pretty sure the Ekos Scheduler will NEVER call the goto *during* an exposure.

The logic is as Hy describes. AFTER an exposure, if the hour angle is greater than the specified flip point, Ekos does a goto to the current RA/DEC. It's the mount (not Ekos) that flips.

Peter, from what you describe, your CEM120 does almost the exact opposite of my CEM40! If my mount is within the "tracking past the meridian" limit and Ekos tells it to slew to its current RA/DEC, nothing happens at all.

On the other hand, you describe your mount as performing a flip regardless of what Ekos is doing. In that case the correct thing for you is to just disable the meridian flip in Ekos and let the mont take care of it.

Remember, all Ekos can do is a goto to the current coordinates. It's the MOUNT (not Ekos) that has to know it's time to flip. Many mounts can track well beyond the meridian. If you have your mount set to track say, an hour past the meridian and Ekos tells it to slew to current RA/DEC 5 minutes or 30 minutes past, nothing will happen.
2 years 4 months ago #77857

Please Log in or Create an account to join the conversation.

  • Posts: 992
  • Thank you received: 155

Replied by Peter Kennett on topic Meridian flip

OK, now that I am waking up - I realize my comment was not accurate at all. That's what I get for commenting whole half asleep. What I experience is NOT the mount messing up my exposures, but my DOME.
I cant figure out why - but it flips when the mount reaches some point near the Meridian. That's why my exposures get destroyed during the flips. EKOS doesn't know that the dome has moved. I don't even know WHY the dome needed to move. It just goes around and stops at the the position. There are several issues with the INDI Nexdome driver and this is just one of them.

Sorry for the confusing posts.

Peter
2 years 4 months ago #77858

Please Log in or Create an account to join the conversation.

  • Posts: 1185
  • Thank you received: 370
I think all the comments describe well how the MF mechanics are. Glad to hear knowledge about this tricky thing spreads :-)

Just one thing about the steps after the slew has happened. Capture re-initiates all things that where in place before the slew happened. If EKOS was guiding before, it will restart guiding. If there was an alignment, it will redo it. I.e. you do not need the scheduler for these post-flip steps to happen, it just automates them.

Not using this automation is simply more error prone. E.g. if you just turn off guiding manually before the flip happened, it won’t be restarted automatically.
2 years 4 months ago #77868

Please Log in or Create an account to join the conversation.

  • Posts: 472
  • Thank you received: 165

Replied by Jarno Paananen on topic Meridian flip


Do you have hour angle selected for pier side in dome slaving parameters or mount? Later should be used if mount reports pier side correctly, hour angle is workaround for that, but causes issues near meridian when mount is on the "wrong" side.
2 years 4 months ago #77873

Please Log in or Create an account to join the conversation.

Time to create page: 0.894 seconds