Meridian Flip failure with The SkyX

I have a Paramount MX+ Mount and I am using SGPro to sequence my captures interfacing with TheSkyX using the ASCOM Telescope Driver for The Sky.
Up until recently I have not needed to do a Meridian Flip. Unfortunately it is usual for PHD2 Guiding to lose the guide star during autofocus as I have specified to turn off guiding for the focussing process. During the recovery process after autofocus, when re-centring, SGPro tells The SkyX to slew to the RA and Dec coordinates of the target. For The SkyX I have set limits beyond the Meridian to avoid the OTA colliding with the pier. For my set up that is within about 1 hr 50 mins of it. If the OTA has started on the west side of the pier (I am in the Southern Hemisphere), but the target has rotated to within 1 hr 50 mins of the Meridian, when it is asked to slew to that target it will automatically flip to the other (east) side of the pier, so that a flip is no longer required for the rest of the night. Up until now there has always been the need for a refocus when the target was within the 1hr 50 min envelope next to the Meridian, which has resulted in the flip occurring.
More recently I have been doing work on the tracking of the Mount and in particular getting a better TPoint model in The SkyX and using Protrack, to greatly improve how well the Mount holds the guide star. I am therefore hoping that when PHD2 Guiding is restarted it will reacquire the guide star and start guiding again. I therefore need the Auto Meridian Flip to work. I have set SGPro to flip 80 mins past the Meridian.
I recently started imaging with the OTA on the west side of the pier and frankly due to very steady temperatures , no autofocus occurred before a Meridian Flip became due. However it failed. Here is a DropBox link to a folder into which I have placed the SGPro log for the run, including a picture showing when the Meridian Flip was initiated and failed. I have also included the PHD2 Guide log for the run.

I would like to understand why the flip failed and what i need to do to get it working.
Thanks,
Niall MacNeill
Wattle Flat,
NSW,
Australia

I have MX and MyT and the good news is that the flip works just fine.

Please check out my YouTube video on SGP/TSX. It explains how to set things up.

I cant really say what happened except to note that there was a failure within the bisque driver itself. Beyond this, we dont have any visibility:

[12/09/20 22:52:17.353][DEBUG][Telescope Thread][SQ;CE;] ASCOM Telescope: Error in Slew : CheckDotNetExceptions ASCOM.SoftwareBisque.Telescope SlewToCoordinates System.NullReferenceException: Object reference not set to an instance of an object.

Thanks Chris, your YouTube tutorial was most useful. I adjusted the settings to match yours, with the following exception. I had my slew limits set to 1.5 hours before or after the Meridian, based on not crashing my gear into the Mount. Based on the information you provided and given I do some narrow band imaging at a 30 min exposure time, I set the slew limits to +/- 45mins and minutes past the meridian flip to 32.
Here is a DropBox link to the settings I have used:

I had another failure however after I changed these settings. When it came to doing the flip, I got a message saying the Sequence was ending. I have included a picture of the relevant time from the log and some of the messages, but am really unable to diagnose what went wrong.

Hi Chris,
Happy New Year.
Can I ask you a couple of questions on your YouTube video " Sequence Generator Pro (part 12) - getting Paramount and SGP to play nicely". As I said, this is a great reference…thank you again.
From 7:55 you go into the Control Settings in SGPro for the Telescope to manage the Meridian Flip and you choose to allow the scope to go 22 mins past the Meridian on the basis that your longest exposures are 20 mins. From what I understand this is because you want to allow an image to be captured at any stage before the OTA passes the Meridian, but not much afterwards. Why not afterwards? You set the software slew limits to 30 mins each side of the Meridian so that the flip will always occur within the boundaries. A friend of mine simply sets the “Minutes past Meridian to Flip” to zero and leaves the “Wait for Meridian” unticked. Mind you he using an EQ Mount. What this means is that if the Mount has not yet reached the Meridian it will start an exposure. He knows his longest exposure of 30 mins does not go beyond his slew limits. If the Mount finishes an exposure after the Meridian has been passed, it performs a Meridian Flip. Is there any reason not to use this second method?
My software slew limits allow me to go +/- 1.5 hrs of the Meridian, without crashing the OTA into the pier. Is there any reason why you wouldn’t allow the scope to get closer to this limit before forcing the flip. So if my longest exposure is 0.5 hours, could I not set the “Minutes past the Meridian to Flip” to say 54 mins (0.9 hours), such that an exposure started at 53 minus past the Meridian would finish with the OTA 83 mins past the Meridian, but still within the 90 minute slew limit?
As I mentioned previously, for my most recent effort and in order to mirror your settings, I set the “Minutes past the Meridian Flip” to 32 mins, since my longest exposure is 30 mins and the slew limit was set to 45 mins past the Meridian. Should that have worked?
You said that the "Wait for the Meridian” dialogue box should be ticked “so it doesn’t accidentally flip with the weights still pointing up”. I don’t understand this. What is the problem with the Meridian Flip starting with the counterweights pointing up as long as the OTA is not in danger of hitting the pier. The mouse-over says “if the time until flip is less than the next exposure then the sequence will wait until after the meridian flip to start another image”. So ticking this box means that it won’t start an image if that exposure won’t finish until after the meridian flip is due, but will flip straight away, with the corollary that if it is not ticked, an exposure will commence that will be interrupted by the need to flip. Is that your understanding?
In terms of PHD2 Guiding I have the Camera set to “SGP API Guider (ASCOM)”, which allows PHD2 to access the guide chip on my SBIG 16803 camera. For the Mount I am, like you, using “ASCOM Telescope Driver for TheSky”. You said you use this too, because you are using a “Pulse Guide” system through the SkyX, but earlier you said you use the “Direct Guide” System? The instructions for PHD2 indeed say that “The ASCOM driver must support the ‘PulseGuide’ interface”. When using “Direct Guide”, I presume this is also supported using “ASCOM Telescope Driver for TheSky”. If not what is the correct configuration to run Direct Guide?
I would be grateful for any help.
Regards, Niall

Thanks Ken,
It has been a bit difficult to understand what is going on. In the latest failure, a Ha image was saved at 23:38:00.766. The next images was to be a 10 min Luminance sub.
23:38:06.433 it checked if a Meridian Flip was needed, then noted the OTA was on the west side of the Mount and that a Flip was not needed.
Hour Angle < Degrees Past to Flip: 3.626 < 8. The 8 degrees corresponds perfectly to the 32 mins past the Meridian that I specified.
The 3.626 corresponds to 14.5 mins past the Meridian so there were 32 - 14.5 = 17.5 mins left to make a 10 min capture. Conclusion no Meridian Flip required. Therefore at 23:38:06.433 there were 17.497 mins left to the Meridian Flip, which was set at 23:55:36.256
But since the next image was to be Luminance a refocus was required. Shouldn’t the Meridian Flip check be done after a refocus like this, since if the refocus took more than 7.5 mins, then the capture couldn’t be completed before a Meridian Flip was due?
23:38:06.444 It noted that a refocus was required since the Temperature had changed by more than the 1 degree I specified. But since changing the filter an autofocus was required anyway.
23:38:06.461 Autofocus commenced
23:43:48.399 Autofocus complete…Delta 5 mins 42 secs.
That leaves just enough time but would it not have been better to do the Meridian Flip check here?
23:44:57.223 Luminance exposure starts
23:55:09.446 Camera read complete…about 26 secs before Meridian Flip due…whew
23:55:21 checks if Meridian Flip is needed and confirms that OTA is on the west side of the pier.
23:55:21.099 “Presenting wait for Meridian Dialogue”…what does that mean?
23:55:32.344 “FlipWait: ending wait due to telescope parked or not tracking…” I can think of no reason why the telescope would have stopped tracking. The slew limits were set for 45 mins, which means the Mount should have keep tracking at the sidereal rate until something like 00:08:36.25
23:55:32.833 Aborting Sequence: User aborted meridian flip.
I recall getting a Dialogue Box saying the Sequence was ending.
Any ideas as to what has happened here?
Thanks & Happy New Year.
Niall

In terms of meridian flip settings - I wait for the meridian, so, to avoid wasting valuable time, I pitch the settings that it only has to wait around for a few minutes. So if it 22 minutes and the longest exposure I ever do is 20 minute… you get the idea. If you try and force a flip before the meridian, the weights end up being higher than the telescope on the other side. I am not even sure if you can do that through ASCOM with a Paramount.

Guiders operate in a number of ways. In the early days, by using physical relays to move the mount motors. While the ST4 interface still exist, most modern mounts accept software based rather than hardware based guiding commands. The generic name is pulse guiding, of which direct guiding is Software Bisque’s take on it. I have used all three with the MX. In practice, I have not seen any meaningful improvement of direct guide over pulse guide, though I understand it operates, under the surface, in a different way. I recommend disabling guiding during slewing or centering, I have seen conflicts between direct guide commands and others, causing software exceptions that then block further guide commands from PHD2.

The Paramount can fail the meridian flip for a number of reasons. Double check the little square box is on the meridian in TSX and try not to flip on the meridian itself. Too many small ambiguities on the meridian. Also check your signage in TSX for your limit settings, one should be negative, the other positive.

The TSX API can also throw out exceptions, especially if it is being bombarded with simultaneous commands. Remembering that TSX can drive many different mounts, the ASCOM driver has to accept them for what they are and it disables some of the ASCOM available methods/properties. It is highly likely that if you take a look at the ASCOM properties, some of the checkboxes will be clear.

Thanks you all for the information. I’ve properly configured the Paramount MX and I’m eaguer to test those meridian flips. I’m pretty sure it will work fine.

As fot TSX, as Chris said, I’ve had problems with DirectGuide and TSX changing the Ascom driver setting on his own, which i think it shouldn’t happen.

Regards

Blockquote In terms of meridian flip settings - I wait for the meridian, so, to avoid wasting valuable time, I pitch the settings that it only has to wait around for a few minutes. So if it 22 minutes and the longest exposure I ever do is 20 minute… you get the idea. If you try and force a flip before the meridian, the weights end up being higher than the telescope on the other side. I am not even sure if you can do that through ASCOM with a Paramount.

Blockquote

Thanks Chris. Does it really matter if you flip before or after the meridian in terms of lost imaging time as long as you are within the slew limits each side? If you flip before the meridian, the OTA will still go across to the eastern side (in my case, western for you) and you will start imaging again as soon as the flip functions such as refocussing, re-centring and guiding set up are performed. Yes the counterweights will be somewhat raised, but that shouldn’t mean any wasted time. The same argument applies if you flip well after the meridian. Do you see what I mean?

Thanks for the summary on direct versus pulse guiding. I have so far set up with pulse guiding without difficulty, but will try Direct Guiding when I can. SGPro does disable guiding during slewing and centring. I also set it to disable it for auto-focussing and that does cause me some issues. During the 5 minutes or so of autofocusing, I have found PHD2 loses the guide star. This forces a recovery, involving re-centring, which costs time. I will try increasing the guide box size in PHD2 and have also taken steps to improve my unguided tracking (TPoint model, Protrack and getting tube rings instead of the dovetail attachment).
Hopefully I can get some clear skies to do some more meridian flip testing to see why it keeps falling over.

Not all mounts implement a flip command (or side of pier) in ASCOM. Some flip automatically with a slew to command. If you let a Paramount track past the meridian and then tell it to slew to the same spot, it will flip. I cannot recall if a Paramount also has a flip command through ASCOM. I guess one way to try would be to write a VB script and see what happens. If a mount does not have a flip command, I do not know of any other way than to wait for it to go past the meridian and issue a new slew.

Thanks Chris, that is interesting. If you issue a slew command before you reach the meridian, but within the slew limits (purple coloured area), the Mount will also flip. It seems that if you slew to an object within the slew limit area, TheSkyX will always move the OTA to the eastern (western in your hemisphere) side of the pier, to minimise the need for a future flip. This means that if you are within the slew limits but before the meridian (purple zone), the starting position will mean the counterweights are up.
Incidentally, one reason I have not really needed to bother with a meridian flip up until now, is that inevitably before the object gets to the meridian, an autofocus is required. Since PHD2 loses the guide star, as I described previously, SGPro needs to go into recovery mode and that means recentring on the object. When SGPro tells TheSkyX to go to the RA/ Dec coordinates, the Mount flips if the object is within the slew limits (purple zone).
I am also unsure how the meridian flip is supposed to work between SGPro and TheSkyX. I would have thought, the easiest way would be for SGPro to simply tell TheSkyX to go to the RA/Dec coordinates. As mentioned above if the OTA is in the purple area (before meridian) or red area (after meridian), the mount will automatically flip.
The way you have it structured the Mount will be in the red zone after the meridian, when the flip is initiated. But what exactly happens then? Does SGPro simply tell TheSkyX to go to slew to the RA/ Dec coordinates or does it actually give a command to flip?
What is a VB script?
Thanks Niall

Thanks for the info - I have never sat down and experimented with TSX in this way.

VB - visual basic… (never used SideOfPier before so don’t know if the following works)

set c = CreateObject(“ASCOM.Utilities.Chooser”)
c.DeviceType=“Telescope”
id = c.Choose("")
set scope = CreateObject(id)
scope.Connected = true
scope.SideOfPier = pierWest
scope.Connected = false

Has anyone attempted to do long UNguided imaging using ProTrack while the counterweights were UP? From what I can tell, T-Point doesn’t take any counterweight-up points, so ProTrack doesn’t have any data to use when the counterweights are up, so tracking isn’t as good. I’ve run multiple unguided tests with weights down, then up, and once the weights go up tracking gets significantly worse. To the point where I can’t do unguided imaging (I normally take 15 minute subs on my ME II with encoders), so I now flip at the meridian.
Eric

Hi again Chris,

Just playing around with settings, I see that the side of the pier to which the OTA slews depends on where the Flip Hour Angle is set. If for example it is set at -1hr, then if you ask TheSkyX to slew to an object which is within an hour of reaching the meridian (ie in the purple zone) it to will go to the east side of the pier for Southern Hemisphere users and the opposition for you guys in the Northern Hemisphere.
I hadn’t realised it but the default must have been set to the extremity of the Mount slew limits to maximise the opportunity for the Mount to slew past the Meridian and then be able to track objects through the Meridian.

Hi Eric, I have my C14 mounted on a dovetail and the flexure in the connection to the Mount is such that I can’t go unguided. Also the TPoint Model precision is limited by the connection so that Protrack does not have an accurate enough model to give me good unguided tracking. I am sourcing tube rings in the hope that either I can go unguided or that the guiding performance will be better. So I I’m sorry but I can’t help you with your question.

Yes - been there, seen that, got the T-shirt. That is why I specifically mention this setting in the YouTube video I posted earlier in this thread. :slight_smile:

Thanks Chris. I am much better informed for that. I still haven’t had an opportunity to get to another meridian flip to see if it works.

Hi Chris,

I haven’t had much in the way of clear skies lately and the object I’ve been imaging has a dearth of guide stars on either side of the meridian. Therefore I haven’t been concerned with getting the meridian flip to work in SGPro.
However, I have recently started imaging NGC 5128 and it is in a nice star field with plenty of guide stars each side of the meridian. It was actually clear enough the other night to require a meridian flip which again failed.
I went back to our discussion and decided to see what I could do to understand why the meridian flips were failing and therefore to get it to work.
A couple of nights ago I finally had a success. The issue was that the Flip Hour Angle on The SkyX has to match that chosen for the meridian flip on SGPro.
For The SkyX we define the slew limits, which is how much time before or after the meridian to avoid hitting the pier. For me this is +/- 1.5 hours. The SkyX allows one to flip to the other side of the Mount by pressing a button on the Telescope Page, but only if the OTA is pointing within the flip envelope i.e. the purple (before) and red (after) meridian zones, defined by said slew limits.
However, the Flip Hour Angle is important to determining to which side of the pier the OTA will move when slewing to an object that is within the flip envelope. If it is set to 0 hours, then if the object is in the purple zone, before the meridian, the OTA will move to the east side of the pier (or the west for me in the Southern Hemisphere). To avoid confusion I am going to reference the relative positions for the Northern Hemisphere. If the object is after the meridian in the red zone, then the OTA will move to the west side of the pier when the slew is initiated.
However the Flip Hour Angle can be used to alter the side to which the OTA moves. For example if it set to + 1 hour, then if the object to which we are slewing is before the meridian or up to 1 hour after it, the OTA will move to the east side of the pier. If it is more than 1 hour after the meridian it will move to the west side.
Conversely if the Flip Hour Angle is set to -1 hour, then the OTA will be on the east side of the pier until up to 1 hour before the meridian and after that it will be on the west side.
It is not difficult to see that if the Flip Hour Angle is not consistent with the hours before or after the meridian set in SGPRo, there is a potential inconsistency between the side of the pier to which The SkyX is moving the Mount and where the SGPro expects it to go to.
A couple of nights ago, I set the Flip Hour Angle to 0 hours and for SGPro I set the flip to take place 0 hours after the meridian. i.e. consistent. The flip worked perfectly.
Last night I set both to + 1 hour and dutifully the flip occurred 1 hour after the meridian.
This can be very useful if for example, you don’t want a meridian flip to take place, but you want to start imaging as soon as possible when the object entered to flip envelope…in my case 1.5 hours before the meridian. In this case I would set the Flip Hour on The SkyX to -1.5 hours and same for SGPro. Then the sequence will be set to start say 1.49 hours before the meridian. At the start SGPro will tell The SkyX to slew to the object. Because of where the Flip Hour angle is set, the OTA will move to the west side of the pier to find the object and it can then stay on that side of the pier all night without the need for a flip. SGPro should report no flip required.
Does this all make sense to you? I didn’t pay enough attention to having a consistency between the meridian flip timing on SGPro and the Flip Hour Angle in SGPro and the resultant inconsistencies led to the failures.
CS Niall

Yes - I think I follow you. I drag the meridian flip line in TSX to the meridian and in SGP, always wait for the flip point, rather than try and go early. In that way, I have not had any conflict. The YouTube video mentions my other settings - with SGP flipping before TSX stops tracking.

Exactly Chris.
The way you have set things up is consistent with not creating a conflict. Since you set the flip to occur 22 mins after the meridian and tell SGPro to wait for the meridian, it will always have gone past the meridian when it starts its last exposure before the flip. When the Mount flips, SGPro is expecting the OTA to move from the east to the west side of the pier. Since, when The SkyX is asked to point to the object you are imaging, it is in the red zone already, it will move the OTA to the west side of the pier. No conflict.
The other alternative to the way you’ve set it up and a simpler method I believe, is to set the “Minutes past the Meridian to Flip” to 0 and only tick “Auto Centre after Meridian Flip”. i.e. you do not really need to tick “Wait for Meridian” and then set a flip time to be just longer than your longest exposure. For my proposed “simpler” set up, SGPro will start an image if it is before the meridian and it will finish it, even if it goes past the meridian. Once that image is finished, the object will be past the meridian and SGPro will initiate a flip. Since the Flip Hour Angle on The SkyX is 0 degrees, as you specified in your tutorial video, this is in the red zone and when asked to point to the object, The SkyX will move the Mount such that the OTA moves to the west side of the pier, as SGPro expects…no conflict.
The only thing one has to ensure is that the slew limit is set such that the OTA can go past the meridian by more than the longest exposure!
Incidentally, it may be worth considering just that scenario i.e that someone may have an exposure time which is greater than the slew limit past the meridian. Imagine a 60 min exposure and 30 minute slew limit for example.
Your set up would not be possible, because you couldn’t set SGPro to flip after the slew limit for obvious reasons. For the set up I’ve suggested, there are circumstances where the Mount will reach the slew limit before an exposure, which was started before the meridian, is complete.
In this case, there are three alternative configurations to avoid an infeasible outcome as discussed above.
Firstly the astro-imager could tick “Wait for Meridian”, where the “Minutes Past Meridian to Flip” is set to 0. In this case, supposing the Mount is 59 min before the meridian, since the exposure will take 60mins and go 1 min past the meridian, SGPro will wait until the meridian and then flip. This means a 59 minute loss of imaging time in the worst case. In the best case an image started 61 mins before the meridian will finish 1 minute before the meridian and just before the flip, with minimal lost time…just the safety factor. The average lost imaging time will be ~30 mins.
The second alternative is to set the “Minutes Past Meridian to Flip” on SGPro, as well as the Flip Hour Angle on the SkyX, to just before the slew limit, say 29 minutes and tick “Wait For Meridian”. In this case if SGPro goes to start an image say 30 mins before the meridian, the 60 minute exposure will take it past the meridian flip time and it will wait. Again a loss of 59 mins. If it starts an image 32 mins before the meridian, which will finish 28 mins past the meridian, there is a loss of 1 minute. In fact alternative two is no better or worse than scenario 1. Both avoid a failure, but result in an average loss of imaging time equal to half the exposure duration.

The third alternative and the best solution is to set the flip to occur before the meridian. In your video you suggest that this is not a desirable situation, but in fact there is no problem with it. In this case you would calculate how many minutes before the meridian the flip needs to occur to avoid a conflict or a significant loss of imaging time. This = Texposure - Tslew limit = 60 - 30 = 30 mins. So the flip needs to occur 30 mins before the meridian. In this case I would set the Flip Hour Angle on The SkyX to -32 mins and the same for the “Minutes Past Meridian to Flip” on SGPro. “Wait for Meridian” should be unticked. (By the way the word “Meridian” in this dialogue box in SGPro is a misnomer . They should say “Meridian Flip Time” as it may not correspond to the meridian at all)
If SGPro arrives at a situation where it is say 33 mins before the meridian, it will start the exposure, which will finish at 27 mins past the meridian. 2 minutes later a flip will be initiated and that constitutes the lost imaging time i.e. again the safety factor. If however, SGPro finishes the last exposure such that it is 31 minutes before the Meridian, it will initiate a Meridian Flip with no loss of imaging time, save that for the flip itself which is inevitable in every scenario.
Chris, do you agree with all this? If you do, it may be worth considering a reworking of your tutorial video. You have a very nice presentation style and I think there would be benefit to firstly simplifying the basic scenario as I suggested and how to manage scenarios where a Meridian Flip at the meridian is not always desirable. If you don’t want to go into such detail, perhaps you can just mention that there are scenarios where the imager may want to flip before or after the meridian and that in such cases they need to ensure that the Flip Hour Angle on The SkyX and the “Minutes Past Meridian to Flip” are the same.