The new driver version is still 0.7 for now, so it is not significant. But if you are running the new version you must see the mount type selection buttons under Mount Info as in the screenshot below. Otherwise, you are running the old one.
For scope without absolute encoders, like celestron, at least one alignment after power up is needed. I suggest to take the first sync point with the scope pointing direction according to some marks on the mount itself. Then you can PS 2 or more point. After these syncs, you can delete the first sync point only, since its accuracy is worst than PS. The alignment subsystem needs at least 1 sync point, otherwise celestial to telescope coordinate conversion goes wrong and the scope makes crazy movements.
Yes, this is the correct behavior.
Great Kevin! Only be sure to run the new driver version.
That’s odd I’m definitely on the old driver then. And I followed your install without any errors. I’m still on the stable version, I was going to switch to nightly. The only other requirement I didn’t check for was the 3rd party drivers. Are they not installed on the stable release?
You touched on another question I wanted to ask. Your coordinates of 0, 0 is that just an example? I’d like to use coordinates of Polaris. I’m always doing a polar align before this and I know I’m roughly pointed there?
I know plate solving by the pole is not recommended. But this is manually sync point so maybe it’s OK.
The driver is only in my github repository that is forked from the official indi-3rdparty library. You can find it here github.com/fabriziop/indi-3rdparty . Follow the build instruction in README. Pay attention: the new driver version is in the "celestronaux_eqtrack" branch, so if you clone the repository on your PC, before building, you must switch from the default branch ("master") to "celestronaux_eqtrack" with the command "git checkout celestronaux_eqtrack" on your PC in the cloned repository.
I changed the version of the new driver to 0.8, so you can immediately check.
My celestron set its internal coords to 0,0 at power on. By celestron convention this means Az = 0 = azimuth pointing North and Alt = 0 = altitude perpendicular to azimuth axis. Since the alignment subsystem of celestronaux needs at least one sync point to work, put your scope into your preferred position and sync with the coordinates corresponding to that position. The alignment subsystem will do the proper coordinates conversions to cope with the celestron initial setting. Using Kstars 3.5.0 Mount Control Box, this input of coordinates is easier: If your mount is equatorial, select HA/DE, if your mount is AltAz, select AZ/AL.
So, I was able to update the driver. Thanks for your help! Here are a few things that I’ve noticed through the alignment process that you have described.
I selected the new EQ option under mount info.
Upon the initial connection the alignment sub system is not enabled. I got a warning message notifying me of this. When I went to the alignment tab, it showed as green, but upon selecting it it was obvious that I just enabled it. So perhaps it should be enabled by default, or not show a green when it’s disabled. I’m not being critical, just want to be thorough.
I found the manual alignment worked best when I used the search option in the mount control. When I search for Polaris and then added the coordinates it seem to work better than when I manually added the coordinates. This might be because I just used 03,89 which I thought would be close enough.
If I tried to slew to a star in Kstars, I found that most of the time the mount would move in wild dramatic ways. If I just use the mount control to slew manually and then plates solve I found this to work better. With only using the simulators I’m not sure how real world my results are.
After adding several sync points/indexes through PS, the way I removed the first manual sync point was to slew back and then select remove current index. If possible it might be better to have a remove first sync point/index option?
When I try to slew to nearby star using kstars, I’m still getting wild movements from the mount, unless I add syncs by slewing using the mount control.
The green light fooled me too. I wasted some time to notice that the alignment subsystem is disabled by default, so I added the warning message. The management of the alignment subsystem settings page is done inside the subsystem itself and I was reluctant to put my hands inside it, at least for now.
It is not necessary to slew back, simply delete the first point. Be sure that current index is zero, since the first point has index = 0 . About adding a specific delete option, the same goes for the alignment page above.
When you slew with Kstars, do you have already done some sync points?
To answer your question, when I slewed in Kstars I only had the first manual sync done through the mount control.
Maybe my process is incorrect. Here is what I'm doing.
Mount setup and pointing roughly at Polaris
Connect Ekos to all equipment
Turn on alignment subsystem
Use Mount Controls in Ekos to enter Polaris coordinates and do a manual sync through Mount Control app (by the way, I love the new options in the app)
Now the Current Set Size is 1 and the mount is reading the same coordinates (in the align tab) as the manual sync
I would then use Kstars to slew to a star close to Polaris (like Kocab) and would like to do a PS and sync (but here is where the crazy slew happens)
If I move the mount with the mount controls (press and hold an arrow to slew) and do a PS and sync this works. I would then repeat this step and add another sync point.
With the Current Set Size now reading 3 (1 manual sync and 2 PS and syncs) I would remove the first manual sync?
Is the Current Set Pointer just listing which sync? It normally reads 0, but I can increment it and select the manual sync and remove it?
You mention that the first sync is 0, but as soon as I do a sync the Current Set Size changes to 1, so shouldn't I select 1 and then choose delete current index?
I would then have 2 PS syncs still saved, I could then add more if needed.
Sorry if this is too much detail. It is a bit confusing to use the terms sync and index. Perhaps they are different, but they seem the same to me?
Sync is the operation that tells to the alignment subsystem: take current scope pointing direction obtained by PS or by manual coordinate insertion, match it with current reading of Az/Alt coming from scope motors, store them into a data base internal to alignment subsystem and use them for the computation of the transformation matrices that converts telescope pointing direction (given by reading motor position) to corresponding celestial coordinates (RA/DEC) and viceversa.
About first sync deletion. I spoke about sync index, but it is called "current set pointer" into alignment tab. The "current set size" tells you how many valid sync points you have done. So, after the first one, it goes to 1. After 2 more PS sync, it goes to 3. Otherwise, the pointer starts from 0, this means that first sync has pointer=0, the second has pointer=1 and so on. To delete the first sync, let current set pointer = 0, select as "action to take" the choice "delete entry at current index", finally click the OK button of "Execution". The pointer sometime is called pointer and sometime is called index.
The wrong thing is the crazy movements after the first sync, that should not happen. I'll investigate it.
I hope you don’t mind me joining this conversation, I am also experimenting with this driver on a NexStar 6SE mount. Sometimes when I go to a target the scope hunts around the target, often for about a minute. Is this the crazy behaviour you are talking about?
BTW I am having zero luck with plate solving but that is probably because I am a novice and don’t understand what I’m doing. Having said that I have some software engineering experience and have recently fixed a bug in the regular Celestron driver so I may be able to help out with this driver.
My scope is off site now and since here there are in force several restrictions due to the pandemic, I am not able to do real tests. So if others have this capability, it is greatly appreciated. Moreover, if other expert eyes look at the code, we may have a faster debug.
About unexpected motions, I have no test for now. Probably, Kevin has seen more. To debug motion, I suggest to do a first sync, for example pointing to south meridian with DEC = 0 (for eq mounts) and then do a goto with the Mount Control Box, changing only one axis at a time.
I actually saw something today that showed some issues with motion control. When I went through the procedure above, I noticed that when I manually moved the mount with motion control, the DEC axis moved correctly. But the RA axis was not moving, or barely moving. I’m not sure if this has anything to do with the AUX driver? But with the crazy motion issues, I thought I’d let you know.
Welcome Ken, and thanks for your work on the driver Fabrizio.
I did some tests with the scope indoor on equatorial mount (Celestron CPC1100). So, I can only make qualitative judgments about pointing directions.
First, I pointed manually the scope to HA=0/DE=0, set the HA/DE values into Mount Control Box to 0/0 and clicked on Sync button.
Then I made several GOTOs using the Mount Control Box and using the Kstars chart. I have not seen strange motions and each pointing direction looks good (qualitatively).
Right now, I have no chance to test equatorial tracking, to see if it has the correct speed, etc.
I had a good indoor session today with NexStar 6SE, plate solving, goto, guiding, focusing etc. All in Alt/Az mode as I don’t have a wedge, (my next 3D printing project)
No problems, one observation, sometimes when manually slewing the mount moves and then stops, another button press and it moves and stops etc. Same if using the hand controller. This may be some setting I don’t know about but most of the time it doesn’t do this.