It looks to be what we have been saying all along. The flip is failing because it is too close to the meridian.

A slew finishes at 00:03:29, the pier side is West so Kstars expects that a slew will be needed soon.
At 00:04:00 the flip slew starts. The hour angle seems to be very close to zero.
Then 4 seconds later the slew finishes and Kstars reports that it failed because the pier side did not change. The system is set to retry in 4 minutes but you stopped it before then.

You need to allow the mount to track further past the meridian At present your limit is 0.02 hours, 1.2 minutes.

One thing though, it looks as if the first slew was to a position that was past the meridian - a positive hour angle - this slew would b expected to have already done the flip. I would experiment by doing a series of manual slews to see at what hour angle the mount will flip.


Do I understand that there is a zone of the sky close to the meridian which your mount cannot enter because the scope will hit the tripod?

You avoid this area by setting the mount Ra axis limits to -5 degrees rather than the default of +20 degrees. This means that the mount will stop tracking 20 minutes before the scope reaches the meridian, at an hour angle of -20 minutes. You then need to wait 40 minutes, until the target is at an hour angle of +20 minutes before a slew to that position works.

If that's correct, that there is a zone close to the meridian which the scope cannot enter, then at present I don't think that Ekos will cope.
There are properties that would help but the underlying code doesn't support it.

The Ha Limit needs to be set so that Ekos stops tracking before the mount hits it's internal stop so this needs to be set to -6 degrees.
The Flip limit needs to be set to a position where the mount can move to the position safely, so this needs to be set to +6 degrees.

In practice the mount will then track to the -6 degree Ha limit and stop. That's where it remains because the scope hour angle, which is used to determine when the flip is done stops changing. what's needed is that in this case the hour angle used changes to being the hour ange of the target and when that is reached, in 48 minutes, the flip will be done, tracking will be restarted and the system can contiune.

It looks as if all that's needed is for the flip management system to change to using the target position rather than the scope position. Maybe only when he Ra limit has been reached. The trouble is that this section of code and the various state machines that manage it are so complex that chaning it is difficult and it is easy for it to break something else.


Your best bet is to try it.

By single axis tracker you mean a mount that will only track the Ra axis, not the Dec axis I suppose? Can this tracker guide in Ra?
If so you should be able to configure PHD2 to guide in Ra only and that should be enough to enable the PHD2 drift polar alignment. If you can't guide at all then PHD2 may be unable to do anything.


not if it doesn't work in the ccd simulator.
It is tricky to implement because it is an offset from where the mount thinks it is pointed. It would need a lot of work to implement something that behaved as a real mount does.

What problem would have having working PEC simulation solve?


The current telescope simulator has mount errors which may be able to generate drift. The MA and ME parameters define the polar align error so should give drift. Try values of a degree or two.


As DerPit says a valid reques but difficult to implement.
What applications handle this? It would be interesting to see wht sort of approach to setting the up and managing the conflicts.


Slewing, solving, guiding works in general all good. One issue I have is that I await my Telecope to be slewed to e.g. the eastern side of the mount when my target already crossed the Meridian to the western side of the Meridian and vice versa. I often need to wait way longer than set in the mount settings until KStars slews the telescote to the right side which costs a lot of patience. It seems to not matter what I set in the mounts hand controller. So my first question is if there is a hidden setting in KStars/Ekos that allows me to set a Meridian-distance-limit to avoid the the mount slews on the wrong side (and risking it to run into the tripod)?

There are no hidden settings in EKOS that I know of.
AIUi you are saying that when you command slews the mount does not end up on the side of pier that you would expect, I suppose that you mean that it you command a slew to a position where the mount should be past the meridian and so in the East (pointing West) state the mount ends up in the West (pointing East) state but still pointing at the object.

This can happen if the mount has a different idea of where the meridian is to Ekos, usually because of a time zone or DST setting problem. INDI isn't very good at managing DST because this isn't in the indi time strucure.

I suggest setting the DST to off in the mount and adjusting the time zone offset appropriately.

You should get a situation where slews to objects cause the pointing state (pier side) to change close to the meridian.

The second - maybe related - problem is that nothing happens when I tell Ekos to do a planned Meridian flip. I set the limit to 3° (as suggested in the tool tip in Ekos, 2° less than the mounts limit). In Ekos I also see nicely the remaining time until the flip. If there is some imaging going on it nicely waits until this is finished and then... nothing. Only a short movement of the scope and the timer tells me again that the flip will come in 3 min. Also, it tries to solve and refind the target (like parts of Ekos think the flip was succesful) but the solver totally screws and corrects in the wron directions with the result it wents further away from the target istead of getting closer. Is there any idea from the community if I may oversaw a setting, a little hook or if there is wrong combination of KStars/Ekos settings vs. the mounts hand controller settings?

This bears out the hypothesis above. the short slew is an indication that the mount is not changing the pointing state. The pier side display in EKOS should bear this out. Ekos detects this and should retry but it the mount meridian is an hour out it will tke an hour of trying before it succeeds and something else may happen first.

Try setting the DST in the mount to off and get the mount working then experiment to see at what hour angle a slew will do a pointing state change. With DST on I think you will find that you need an hour angle of more than +1 hour before the mount flips.

The EKOS munt tab should give useful information, the hour angle and the pier side, which should help with this.


Not sure exactly what you are trying to achieve but there are lots of INDI scope drivers which will return RA, DEC and Pier side.

And the base INDI dome driver, on which a driver for real dome hardware is built, has code to get this data from the scope driver and do the conversion to dome position.


If a filter wheel needs movement in one direction to get accurate positioning that's fine, but that should be resolved by the low level filter wheel control software, not dumped on the high level software.

INDI can probably fix this but it should not need to do so. It looks like a problem induced by the incompetent ZWO developers. Wonder what else they have got wrong.


Chris Rowland replied to the topic 'Celestron CGX-L Park Position' in the forum. 2 months ago

There's a set park position option in the mount driver. Move the mount to the position yu want to park in, then select set park to current position. I'm not in front of a system so can't give chapter and verse.
Once done SAVE the config and your new park position should be remembered.


I've seen so many people moaning that doing this sort of set up and testing has wasted hours of precious imaging time...

The mount property that defines pier side is actually the declination axis position. For one pier side the axis position is in the range -90 to 0 to +90 and in the other +90 to 180 to 270/-90.
If you are at a dec of -30 and do a pier flip the hour angle axis moves through 180 degrees and the dec axis from -30 to +90 to 210 for the Northern hemisphere or -30 to -90/270 to 210 in the south. The mount alignment model sorts all this out to give you Ra and Dec.

You can't use Ha because if you track through the zenith the Ha goes from -ve to 0 to +ve (or just under +12h to +12/-12h to just over -12h) The pier side does not change. Really pier side is a poor name, pointing state would be better.

Some mounts also change "pier side" around + or - 6h hour angle, this makes things OK to the N or S but very worng if you move between the two. There's one driver that uses Ha for the -4h to -9h and +4h to +9h range, the mount reported pier side close to 0 and +-12h except for declinations close to 90 where it reports unknown.