×

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

Bi-monthly release with minor bug fixes and improvements

Astro-Physics Experimental driver not working properly in KStars 3.4.2

  • Posts: 77
  • Thank you received: 16
I recently installed KStars 3.4.2 on an Raspberry Pi 4 with Raspbian OS (32 bit). I have been using the Astro-Physics Experimental driver successfully for many months, however in this latest release the pointing and parking of the mount is extremely off and is wholly unusable in the current state. I have attached the log file of a test I just performed.

I have an AP 900 mount with a GTOCP3 controller. Initially when I first connect to the mount and tell the driver to connect with the mount set to Park 3 (Counter weights down and scope pointing directly North as described in the AP manual). The scope marker shows the telescope to be pointing at the South Poler region of sky, not near the north pole as it should and used to do (I am in the Northern hemisphere). See below:



I immediately disconnect from the mount power the mount down and back on, then reconnect and then it will show the proper marker as below:



At this point I will command to Goto a star such as here in Serpens:



The mount performs the Goto and is pointing in an entirely different portion of the sky, about 60 to 90 degrees off in the right ascension axis. My mount hemisphere switch is set properly to North and always has been. My site data and time are correct. The time on the Pi 4 is correct as is the offset for eastern US time zone.

When I test with an older version of KStars (3.3.7) on a Pi 3 B+ the pointing and parking of the mount work as it should and has until I updated. As reference I have performed extensive testing while the driver was being improved and updated a couple of years back by Jasem Mutaq and Michael Fulbright. Micheal did extensive work to get the parking and other features working correctly and reliably and he and I and others were able to confirm the proper operation of the driver.

I have searched the forum and found that the driver was indeed changed for a single use case in the southern hemisphere. See these threads:
www.indilib.org/forum/mounts/4833-advice...-driver.html?start=0
indilib.org/forum/development/6708-astro...xperimental-cpp.html

In speculation it seems that this change may be why the driver no longer functions properly. In the very least there should have been more than one person (several testers and more is better) testing this driver before it was committed to the release. There were several of us testing this driver while being updated a couple years back to confirm proper operation and it has been working correctly until now.

Can we get some other users of the Astro-Physics mounts with at least a GTOCP3 or newer controller to test this latest driver, including those in the Northern Hemisphere? Until this driver is restored to function, I (and I am sure others) are unable to use our equipment.
3 years 9 months ago #56041
Attachments:

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

  • Posts: 59
  • Thank you received: 19
Can you please tell me which version you are using?

I made the driver more verbose and I'm missing some DEBUG entries in the log file, e.g.
[SCOPE] CMD <:GS#>
[SCOPE] RES <10:02:38.7>

And I found in your log file the line

calcParkPosition: parkPos=3 parkAlt=XX.XXXX parkAz=0.000000 " (removed your latitude)

but parkAz should be 0.1, that's the source line from INDI repo:

*parkAz = LocationN[LOCATION_LATITUDE].value > 0 ? 0.1 : 179.9;

I attached a screen shot of the main control tab of the current INDI github version which shows at the bottom the to park to and the HA/Dec fields.


Kind regards, bye, wildi
3 years 9 months ago #56053
Attachments:

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

  • Posts: 77
  • Thank you received: 16
Hi wildi,

Thanks for the reply and information. Indeed it does seem I am using a different version of the driver. Here is the screen shot of the main control tab on my system:



There are definitely differences in the driver tab. I used this deb repo www.indilib.org/download/raspberry-pi.html to build my system before and maybe something didn't install right. I will rebuild my Pi4 environment, this time using the script file from Robert Lancaster which downloads and compiles KStars/INDI from the GIT. This will take a few hours, after which I will run a test on the system and let you know what I find.

Thanks!
3 years 9 months ago #56055
Attachments:

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

  • Posts: 59
  • Thank you received: 19
Hello

the status of the driver is still experimental and after a break of two weeks I'm now testing it with the GTOCP2 version of the controller.

If you use the current INDI repo be still cautious with a mount fully equipped. I suggest to use park 2 to unpark from and make a small slew from HA=-6, dec=0 to HA=-6 dec +5 deg. If that's ok try to initiate a meridian flip.

Hm, if you used an older version, this driver has not been updated for six years, what went wrong? I did not see any abnormality in the log file you posted.

I 'll go on tomorrow Friday because I'm not yet satisfied with the current version. I'll provide a bash script that fetches and compiles my bleeding edge repo if you agree to spend some time on testing.

Kind regards, bye, wildi
3 years 9 months ago #56062

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

  • Posts: 77
  • Thank you received: 16
Hi wildi,

There is a separate driver for the GTOCP2 because it does not have many of the advanced commands in the GTOCP3 and GTOCP4. This is why the experimental driver was made in order to handle the CP3 and newer control box. So you may want to try the GTOCP2 driver on your CP2 control box.

Yes, it is strange that I am having such issues now on the older driver which has been working well for me until I updated to KStars/Ekos 3.4.2. As mentioned in my first post the mount works fine with the Astro-Physics Experimental driver in EKOS 3.3.7 that I have on a Pi3B +.

I have rebuilt the Pi 4 using the latest script from Robert Lancaster and have run some simple tests. I have attached the log file from that test. The driver appears to be your latest driver as I chose to pull the latest master from his script.

The test is done indoors as I have the mount off the pier for some basic maintenance. I started with the mount in the Park 3 position and set the INDI option to unpark from Park3. The mount in Park 3 position:



I then connected to EKOS/INDI and when connected I also set the "Park to" option as Park 3. I also saved these settings in the INDI configuration file, however the driver seems to not read this correctly.

I then did a small Goto to the star kCephie which was just to the right and slightly down from Polaris at the time of testing. While viewing the mount from behind and looking North the mount should have moved the DEC axis slightly right a few degrees and the RA axis should have also moved the scope slightly right a few degrees, however this is not what happened and here is how it looked when finished with the Goto:



The DEC axis appears to have moved correctly while the RA Axis has moved way off to the left.

At this point I commanded the mount to park and here is where it ended:



It should have returned to the Park 3 position as that is where I set it to return to, however as you can see it did not.

Hopefully the log file will help to reveal something useful along with this information. I am happy to continue to help test the driver so we can get it all debugged and working.

Thanks for your help!
3 years 9 months ago #56111
Attachments:

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

  • Posts: 59
  • Thank you received: 19
Hello LinuxUser

thanks a lot for the docu and the log file. There is still something wrong with the park 3 position:

[2020-06-24T22:20:21.135 EDT INFO ] "[INFO] calcParkPosition: parkPos=3 parkAlt=39.261200 parkAz=0.000000 "

parkAz should be 0.1 :

*parkAlt = fabs(LocationN[LOCATION_LATITUDE].value);
*parkAz = LocationN[LOCATION_LATITUDE].value > 0 ? 0.1 : 179.9;


GTOCP2,3,4: I'm currently working on the integration of these controllers into one driver. The infrastructure for that was already available in the experimental driver and I'm using it now to steer my GTOCP2. Yesterday I un- and parked from park 2 and the meridian flip occurred approximately at the HA=0. I think I'll need about two weeks to complete that.

Kind regards, bye, wildi
Last edit: 3 years 9 months ago by Markus Wildi. Reason: wrong interpretation of the log file
3 years 9 months ago #56114

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

  • Posts: 311
  • Thank you received: 42
I am running a released version. Kstars 3.4.2, (not on a Raspberry) this matches up with the driver in indi 1.8.5
root@:~# apt-cache show indi-bin | grep -i version
Version: 1.8.5~202004252003~ubuntu18.04.1

It seems this should be the same EKos/INDI source file combination you were using in the original post.
A quick test on a gtocp4 mount from park position 2 did not indicate a problem.
3 years 9 months ago #56127

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

  • Posts: 59
  • Thank you received: 19
Hello wotalota

can you post the log file with all debug options set just for the sequence unpark from park 2, park to park 2? Alternatively you can send it directly to This email address is being protected from spambots. You need JavaScript enabled to view it..

Kind regards, bye, wildi
3 years 9 months ago #56130

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

  • Posts: 311
  • Thank you received: 42
Wildi,

It starts off unparked and tracking, so I first parked it then went through the unpark/park cycle.

/Tom
3 years 9 months ago #56132
Attachments:

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

  • Posts: 59
  • Thank you received: 19
Hello Tom

thanks for the log. This is the version before I made changes and is known to work at longitudes West of Greenwich but fails badly on the Eastern Hemisphere. I recommend not to change to a newer version until "experimental" has been removed from the driver's name.

Kind regards, bye, wildi
3 years 8 months ago #56168

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

  • Posts: 77
  • Thank you received: 16
Wildi,

After two days and multiple hours of building test environments and running many tests, I can say that the Astro-Physics Experimental driver currently in the release INDI 1.8.5 (the previous Experimental driver) as well as pulled from GIT (your driver) do not work here in the western hemisphere. They both fail badly. The Experimental driver does work in version 3.3.7 of KStars (from last fall I believe). For example, using your driver after I connect to and unpark the mount from Park 3, then try a Goto Just over to Kocab (which is to the right of Polaris a few degrees and barely lower at the time of the test) the mount RA Axis spins about 180 degrees with the counter weights on top where the telescope should be (which is way off) while the DEC axis moves just a few degrees in the proper direction. I have no logs files from this recent test as it seems the log files were not recorded even though I did have them set to record.

In the intervening months of bad weather we have had here were I could not use my equipment at all it seems I have lost the only good working software environment I had (except for the older Pi 3 I have with 3.3.7). In any case since setting up a Pi 3 in 2016 to use the KStars/INDI software which I really like, the single biggest problem I have had from the start has been a proper driver for my Astro-Physics mount. I have lost far too much time to this problem, notwithstanding the success I was able to have.

In any case, despite my exhaustion with this problem I would still be happy to help test your driver with high hopes it might lead to something good. It does not help that Astro-Physics has not been very helpful in the development of this driver in the past. They are awesome in any other regard to their quality of mount and service. They just will not share any further information about the mount command set other than what has been published up to the CP3 controller. This is unfortunate.

So Wildi, thanks for your help and I do hope you are able to get this driver all sorted out.
Last edit: 3 years 8 months ago by Midwest Astronomer.
3 years 8 months ago #56185

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

  • Posts: 59
  • Thank you received: 19
Hello LinuxUser

I appreciate your work and your report. It makes it not easier at first sight since there are now at least two contradicting reports about the driver's state. Anyway, I found out what goes wrong in the current github version and corrected it. Since un/park is the problem I'm testing it at 7.5 E and 7.5 W and once I'm convinced myself I'll ask you and others to look at it.

I learnt that in 2006 too that AP is reluctant to answer questions but this, my, current problem is not theirs.

Since I do not need clear skies to test I'm still optimistic to have things done within two weeks about.

Kind regards, bye, wildi
3 years 8 months ago #56227

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

Time to create page: 0.301 seconds