1) The hand controller was originally disconnected as suggested in the AUX documentation. I only connected the hand controller for testing. The hand controller is also needed during testing using the serial/RS232 interface.
2) Yes. When I first connect the AUX driver, I select the "Mount Info" tab and change the "Mount Type" to "Equatorial". The default is AltAz.
3) Yes. The "Alignment Subsystem Active" button and the "Inbuilt Math Plugin" button is green and already active after first connecting. To confirm the alignment system is active, I tried toggling the button off and then back on. I also tried the "Nearest Math Plugin".
4) After first connecting and changing to Equatorial, the mount coordinates show as 00 00 00 RA and -00 00 00 DEC.
5) After the first sync, the RA and DEC show the correct values of the star selected. This is correct for both manual syncing and plate solving.
6) The motion control is correct. i.e. "Right" moves the RA counterclockwise and "Up" moves the Dec counterclockwise. This is the same as if I was using only the hand controller without Ekos. This motion is also the same if using the Celestron SkyPortal app.
I've been trying today 4 sync sessions with my evolution mount (all started from alt:0;az:0 and based on plate solving) 3 first I was using "Nearest Alignment" and sorry to say, but was quite bad
first point error about ~20000 arcsec
While using "slew to target" It aborts after few tries at around 130 arcsec error.
Fourth session was using built-in plugging and It was:
first point error about ~20000 arcsec
second 3000 arcsec
third 600 arcsec
And while using "slew to target" It aborts after few tries at around 60 arcsec error.
My issue is still not resolved, You have to use this error prone method with first point directing alt0:az0. If you miss this point by a few degs, your whole calibration process is working bad
And I just run into something odd - after writing last message, pressed again "slew to target" error jumped from 60 to 400arcsec (ok some time passed by) but it goes into loop moving to the same place and plate solving same place, but on kstars map it was moving to right place, and after plate solve jumped into real bad place.
In daytime like today I can point the moon using my muscles, and tell SkyPortal that I am pointing it or just use my NexStar hand controller to do the same. Just one calibrating point with proper leveled mount is enough to follow the moon for hours. No North Alt 0 requirements not any, just point to the moon and follow it. Not a chance with Indi_celestron_aux
I have a dream - to take out my gear, connect to it from my worm home (just like using astroberry with indi server) - press one button (auto align to sky using mount, camera artificial horizon and plate solve) and just enjoy the views....
I started to debug my issues with the indi_celestron_aux driver and made a major breakthrough last evening.
Please forgive my coding methods, I have been familiarizing myself both with the INDI framework and development under Linux.
To summarize the issue, it appears the indi_celestron_aux driver never initializes the motor controllers using the MC_SET_POSITION command. Both the ALT and AZM controllers have no known values and thus no home position (i.e. the mystery slew limits in RA).
For example, I added a "zero" button to the indi_celestron_aux driver. This allows me to initialize the controllers with a home position after startup.
In my early investigation, I found it easy to modify Rick Bassham's driver to support the SkyQ WiFi module without the need for a hand controller. Since my mount does not support automatic alignment (MC_LEVEL_START and MC_LEVEL_DONE), I also modified the code to manually zero the mount at startup.
Testing the mount indoors with the modified indi-celestron-cgx driver, all my primary issues are fixed. After the initial startup, I can fully control the mount through WiFi with no hand controller. I can also successfully goto any object using KStars.
I am hoping to test outdoors in the next few days; my local weather has not been cooperating.
are you using a CPC mount or a cgx. I have a CPC wedge mounted and would like to use the aux driver, but found that the lack of a home starting position was a pain. Your mods seem to fix it....so wanted to confirm I could use it with my mount..
Most interesting! I have a CGE-pro, also with home sensors. I am, however, not very familiar with the different Celestron mounts. Would your changes to the driver be helpful also for this mount, the CGE-pro? I'd be happy to test but I am no developer.
Celestron C11, Skywatcher 100 ED Pro
Atik 383L+, ASI294
Great development progress Craig! It would be great if this driver can also support the Celestron Focuser as well (right now, only Celestron GPS driver does). But at any rate, what remains pending in the current driver?
I am scheduling a debug session with Maciek for next week.
We will try to decide if the problem is indeed marker positions.
It well may be that it works in AltAz by accident of zeroed encoders
at Az=0 Alt=0 on power on. On the other hand it should not matter
since this is a simple rotation which the alignment procedures should
take care of.
Jasem: can you confirm that the arbitrary rotation in both axis should
be handled by alignment?