×

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

Bi-monthly release with minor bug fixes and improvements

Driver OnStep (LX200 like) for INDI

  • Posts: 257
  • Thank you received: 22
test 1: I set up three stars in the same quadrant well up in the sky. since the main camera wasn't focused yet the plate solver got a dark image and I stopped it, switched to the guide camera (asi120cs on a converted 8x50 finder) and started the solver again. After getting an image it almost immediately went to the next star, which was on screen, then took off for the third...then ended without the image showing the third star. Congrats! Nothing went crazy! :D The whole thing took about a minute to complete, making me wonder if it actually waited for plate solving? I have the data downloaded but it reported one missing and went to get it online. This never happens fast.

Resetting for a second run before bed and a 6am start:
I focused the main camera and moved to m39 which was a near miss on main (my guider is never centered) to move off the chosen stars: Deneb, Altair, and Vega and tried again using the asi178/os6in newt @600mm ..and forgot to turn off the focus module(doh!)
As you can see( I hope)
File Attachment:

it had more trouble solving this time... much more realistic. I noticed I also forgot to clear the previous run so it may have been influencing things. At this point it says slewing but isn't moving... I tapped west and it stopped slewing and took another shot.(pretty star). it's not moving to the next now. Aborting procedure.

File Attachment:



I then slewed to the nearby Veil nebula and plate solved. I got the same error. no error reported but would not stop slewing, and when I stopped it, resumed solving. Weird. --it's what I do. :)
Looks like it works procedurally, as for the never ending slew, it doesn't happen when I do a regular slew to target so it may be related to either the platesolver or the interaction with OnStep...or my mistakes, which are good for uncovering oddities others are likely to find.

I'd like to do more but I ran out of time. Hope this helps. more below pushed down as I made notes. Gnight folks.

Noting unrelated error: when trying to save anything to disk Kstars crashes to desktop. I've seen this before and will have to narrow this down and make an issue.




File Attachment:



side note: I used last session's speed adjustments arcsec/sec to calculate the ratio between my estimated gear ratio and the needed one and entered it when I changed the firmware. I just took a 60second photo with round stars and no adjustments from start. Woot!

File Attachment:

Last edit: 5 years 8 months ago by Ray Wells. Reason: pic out of place
5 years 8 months ago #27573

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

Time to chime in to the development process.

INDI drivers should not strive to do too much outside of their scope. Things like clocks are handled by the client. You cannot assume what they want. Now, you can automatically on startup set the time to the system time, that's OK. But once you receive updateTime(..) from an INDI client, then you should set that. There is nothing preventing you technically from ignoring client requests, but that's bad practice. If the user specifies a GPS device, then INDI::Telescope will update the time from it...etc.

As for the crash when to save stuff, I've only seen this reported on Gnome and still has no idea why.
5 years 8 months ago #27589

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

  • Posts: 257
  • Thank you received: 22
@knro Agree on the scope creep issues...set time is working as I described it btw. In the case of OnStep integration, they are a very actively developing system as well so we've mostly been reacting to and working with new toys in development over there. The latest did some interesting stuff last night, running the model routine without much input and updating the arduino. I need more time with it to gauge results however due to operator error. As for the crash I've seen that several times now in different places and it's pretty much anytime I open a dialog. recovery is simple and ekos stays hooked to indi and doesn't crash the connection to server. I'll to open that as an issue at git and tackle it there to keep from muddying the Onstep thread with it...(as if I don't muddy it enough already. B)
Last edit: 5 years 8 months ago by Ray Wells.
5 years 8 months ago #27590

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

  • Posts: 322
  • Thank you received: 31

OnStep for STM32 requires than an RTC be used (because the module has an EEPROM chip, which the STM32 lacks). Remembering the date/time/latitude/longitude/UTC offset is a nice side effect.

An RTC (either DS3231 I2C or DS3234 SPI) can be added to most other OnStep controllers.
The following user(s) said Thank You: Ray Wells
5 years 8 months ago #27602

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

  • Posts: 322
  • Thank you received: 31

Was this using James' code in INDI, or my code in OnStep?

Can you try my code in OnStep? Just download this .zip and use it instead of Alpha.

github.com/kbahey/OnStep/archive/plate-solve-align.zip

Step by step test instructions here (they worked indoor for me)

indilib.org/forum/wish-list/1406-driver-...html?start=216#27466
5 years 8 months ago #27604

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

  • Posts: 257
  • Thank you received: 22
@kbahey: Yep! It was with your firmware and running regular indi/ekos routines. James' stuff was still in there but wasn't accessed. I wasn't sure if it worked the first time, then I broke it the second time when the solver stopped due to forgetting to turn off the focuser frames. <-- I kinda wish that thing would just turn off if you accessed the camera in use on another tab. I tried to be as descriptive as possible while doing it. one reason I try to journal and then go back and edit. --Also your instructions were spot on.


@avarakin Not sure about gps, might be a good question to ask at / the OnStep group
Last edit: 5 years 8 months ago by Ray Wells.
5 years 8 months ago #27605

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

  • Posts: 322
  • Thank you received: 31

Your report on the testing contained a lot of details, which is generally good, but I got lost in all the tangents, so can't locate a definitive conclusion within it.

Which version of KStars and INDI are you using? I am using the latest stable from Jasem's Launchpad.

When I tested it in doors, with a CCD simulator, it worked flawlessly. Did you check the Align status (number of stars request vs number of star aligned) after it was all over.

The instant plate solves are a problem, it takes 10 to 14 seconds to solve a plate (but that is on an older laptop).
5 years 8 months ago #27607

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

  • Posts: 257
  • Thank you received: 22
I didn't come to a clear conclusion either, other than it ran hands off. Frankly, I just ran out of time. It worked procedurally except possibly skipping solves the first time. Something went wrong with the slewing never completing the second time, and I couldn't fix that without starting over.
I also didn't think to look at the status page. I'll do that next time. Summer short nights make it tough, especially with a 6am start.
I did learn that you may need to clear out old solves when you restart it with a different camera. This could have caused the error. New procedures often lead you down a rabbit hole the first few times. With any luck I'll get another hour to try tonight. --and hopefully will remember to check focus on the main camera before I start it. ..and turn off the focus framing before starting.

....aaand clouds. 7-20-18
Last edit: 5 years 8 months ago by Ray Wells. Reason: clouds
5 years 8 months ago #27608

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

  • Posts: 452
  • Thank you received: 71
Hi guys,

I was out of the game for a while and muust admit I am lost when trying to cathc up the stories.
As Khalid already mentionend versions of things are important.
  • Onstep Firmware is moving very fast and has different branches
  • OnStep Driver has version 1.3 Stable and 1.4 which is the development version that is daily synced with Indi
  • Kstars as well is moving everyday
  • Indi , should be the latest git version

I have a concern with latest James Lan implementation.
When connecting to firmware I have a message that OnStep crashed
Kstars Dialog "KStars detected INDI driver indi_lx200_OnStep crashed. Please check INDI server log in the Device Manager."
and the logs
"2018-07-22T09:29:04: Driver indi_lx200_OnStep: *** stack smashing detected ***: <unknown> terminated"*

@James do you notice that?
all except Firmware(1.4b) , latest build
5 years 8 months ago #27681

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

  • Posts: 174
  • Thank you received: 27
I have the same issue with the latest build
5 years 8 months ago #27691

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

  • Posts: 452
  • Thank you received: 71
@avarakin,

I have no clue when this started, do you know the lasted working release?
This could help to search which change is bringing this error.

then we could roll back.

PS: I checked a checout on commit 4d206f3e8058f5f4d9e2beae2d6cf0882ca82e45 and it works, so the bug must come from one of the upstream merge I did since this commit

regards
Last edit: 5 years 8 months ago by Alain Zwingelstein.
5 years 8 months ago #27702

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

  • Posts: 174
  • Thank you received: 27
Around one week ago it was good. BTW, I also had an issue with compiling some other 3rd party driver, I think it was apogee.
We can probably try to use git bisect command which finds a bug using binary search:
git-scm.com/docs/git-bisect

Alex
The following user(s) said Thank You: Alain Zwingelstein
5 years 8 months ago #27704

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

Time to create page: 0.572 seconds