Hi!

This is an old issue - it's been reported before and I thought it was solved. It concerns the Linear 1-pass, where the focuser strangely settles on a different value than the best value. This is now on a totally fresh Stellarmate image, 1.7.6, 64-bit. First time I try I get this interesting result, shown in the attached image. I did not have logs turned on, can do that next time if needed. But is there a known solution?



Magnus

Is it possibly as simple as the AF Overscan being way too large?

Hi!

OK, I could test that. It is the default value.

However, I do not see how that could result in the value for best focus being changed from 737 in the curve, to the identified value of 488 - way off the measurement....

Magnus

Hi Magnus,

Where this type of issue happened in the past there was a problem with the focuser moving from the last position in the sweep (position 15 in this case) to the solution (737). The reason was some conflicts with backlash.

You have driver backlash set to 0 and Overscan set to 100. So what should have happened is after position 15 (which looks to be about 460) the focuser should have moved outward by 377 to position 837 then back inward by 100 to 737.

So what focuser are you using and what backlash does it have?

Since you have driver backlash disabled there shouldn't be any conflict there. That leaves AF Overscan conflicting with any backlash management your focuser unit may have, for example, if the focuser itself is managing backlash outside of Ekos. If this isn't the case here then something else must be going wrong.

If you get a log file with verbose settings on Focus and Drivers and Indi I'll take a look.

Hi!

I'm using a DIY MyFocuserPro2, controlling a Moonlite focuser. No backlash control outside of Ekos and I normally do not use backlash control, relying on the Linear to avoid it by always reaching focus from the same direction.

However, I've had a series of issues with this R-Pi now (like hanging on boot) and I've re-flashed the SD-card and reinstalled it all. And to my surprise, last night Linear 1-pass worked fine!! I'll keep testing but the next few nights promises to be cloudy...

If I get it right, the AF Overscan refers to how far the focuser overshoots when moving out, then to reverse direction and moving in again. Is that correct?

Magnus

Hi Magnus,

Yes, your description of AF Overscan is correct. So as long as your actual backlash is <= 100 your settings should be fine. If you keep the verbose logging on for a while then if it happens again we'll have something to work on. Just be careful it doesn't eat up all your disk space.

Regards,
John.

Hi!

I'm back with this issue. Updated SM to latest stable. And I try linear 1 pass, same problems as before: it settles on a different position than best focus. See screenshot below.


What is happening here? Log attached, this happened just at the end of the log. Anther identical issue a few minutes before. There are some lines about failing to reach positon that I don't get, not sure if it is related to this. Grateful for any advice on this.

Magnus

Hi Magnus,

The is Ekos version 3.6.4 not the latest 3.6.5

Here's the problem...

The focuser is at the last point on the inward sweep (280) and solves the curve (solution=452.878 or 453)...
[2023-06-06T23:09:18.687 CEST DEBG ][ org.kde.kstars.ekos.focus] - "Linear: 1stPass solution @ 280: pos 452.878 val 2.38224, min measurement 2.37725"

It tries to move the focuser to 453+50 (overscan) = 503 by moving outwards by 223
[2023-06-06T23:09:18.725 CEST DEBG ][ org.kde.kstars.ekos.focus] - "Linear: extending outward movement by overscan 50"
[2023-06-06T23:09:18.726 CEST INFO ][ org.kde.kstars.ekos.focus] - "Focusing outward by 223 steps..."
[2023-06-06T23:09:18.771 CEST DEBG ][ org.kde.kstars.indi] - MyFocuserPro2 : "[DEBUG] CMD <:01#> "
[2023-06-06T23:09:18.821 CEST DEBG ][ org.kde.kstars.indi] - MyFocuserPro2 : "[DEBUG] RES <I0#> "
[2023-06-06T23:09:18.822 CEST DEBG ][ org.kde.kstars.indi] - MyFocuserPro2 : "[DEBUG] CMD <:05503#> "
[2023-06-06T23:09:18.825 CEST INFO ][ org.kde.kstars.indi] - MyFocuserPro2 : "[INFO] Focuser is moving to position 503 "

But the focuser doesn't move...
[2023-06-06T23:09:18.825 CEST DEBG ][ org.kde.kstars.ekos.focus] - Abs Focuser still at 280 . State: Busy
[2023-06-06T23:09:18.825 CEST DEBG ][ org.kde.kstars.ekos.focus] - Restarting focus motion timer. State: Busy

Then the problem...
[2023-06-06T23:09:19.289 CEST DEBG ][ org.kde.kstars.indi] - MyFocuserPro2 : "[DEBUG] CMD <:01#> "
[2023-06-06T23:09:19.351 CEST DEBG ][ org.kde.kstars.indi] - MyFocuserPro2 : "[DEBUG] RES <I0#> "
[2023-06-06T23:09:19.352 CEST DEBG ][ org.kde.kstars.ekos.focus] - Abs Focuser still at 280 . State: Ok

That "state: Ok" is the problem. The focuser hasn't moved but reports Ok - this is wrong

Ekos then unwinds backlash by 50 moving you from 280 to 230.
[2023-06-06T23:09:19.352 CEST DEBG ][ org.kde.kstars.ekos.focus] - "Undoing backlash extension. Moving back in by 50"
[2023-06-06T23:09:19.353 CEST INFO ][ org.kde.kstars.ekos.focus] - "Focusing inward by 50 steps..."
[2023-06-06T23:09:19.360 CEST INFO ][ org.kde.kstars.indi] - MyFocuserPro2 : "[INFO] Focuser reached requested position. "
[2023-06-06T23:09:19.437 CEST DEBG ][ org.kde.kstars.indi] - MyFocuserPro2 : "[DEBUG] CMD <:01#> "
[2023-06-06T23:09:19.445 CEST DEBG ][ org.kde.kstars.indi] - MyFocuserPro2 : "[DEBUG] RES <I0#> "
[2023-06-06T23:09:19.445 CEST DEBG ][ org.kde.kstars.indi] - MyFocuserPro2 : "[DEBUG] CMD <:05230#> "
[2023-06-06T23:09:19.450 CEST INFO ][ org.kde.kstars.indi] - MyFocuserPro2 : "[INFO] Focuser is moving to position 230 "
[2023-06-06T23:09:19.451 CEST DEBG ][ org.kde.kstars.ekos.focus] - Abs Focuser still at 280 . State: Busy
[2023-06-06T23:09:19.451 CEST DEBG ][ org.kde.kstars.ekos.focus] - Restarting focus motion timer. State: Busy
[2023-06-06T23:09:20.352 CEST DEBG ][ org.kde.kstars.indi] - MyFocuserPro2 : "[DEBUG] CMD <:00#> "
[2023-06-06T23:09:20.403 CEST DEBG ][ org.kde.kstars.indi] - MyFocuserPro2 : "[DEBUG] RES <P230#> "
[2023-06-06T23:09:20.404 CEST DEBG ][ org.kde.kstars.ekos.focus] - Abs Focuser position changed to 230 . State: Busy
[2023-06-06T23:09:20.404 CEST DEBG ][ org.kde.kstars.ekos.focus] - Restarting focus motion timer. State: Busy
[2023-06-06T23:09:20.404 CEST DEBG ][ org.kde.kstars.indi] - MyFocuserPro2 : "[DEBUG] CMD <:01#> "
[2023-06-06T23:09:20.453 CEST DEBG ][ org.kde.kstars.indi] - MyFocuserPro2 : "[DEBUG] RES <I0#> "
[2023-06-06T23:09:20.454 CEST DEBG ][ org.kde.kstars.ekos.focus] - Abs Focuser still at 230 . State: Ok

So the question is why does the myFocuserPro2 focuser fail to move from 280 to 503. And having failed to move it reports "Ok" as if it has sucessfully moved.

To make progress we'll need input from the myFocuserPro2 folks. If you reach out to them and if they are happy to discuss this, I'll be happy to join in and give my input either on a call on via email, etc.

The "strangeness" has a strong hold on me. 
I was previously affected; the issue vanished with the new backlash compensation technique introduced in KStars 3.6.4, but it has suddenly reappeared in 3.6.5.

I'm running the most recent stable release of Indi and KStars. However, I just discovered that the myfocuserpro2 firmware has been significantly updated (now ver. 328 while I'm running 324-something). 
I'm now rather convinced that it boils down to some sort of issue with the myfocuser2 firmware. The good news is that with the new routine to calculate filter offset, it is rather easy to check if the error is present or not by performing a series of focuser operations and checking the final movement value.

I intend to check to see if the "phenomenon" is still present with the latest firmware release of myfocuser.

Hi!

Yes, I believe it is a question of the firmware in the focuser controller. I'm in the process of updating mine and testing - however, so far I've not been able to update it (Arduino issues, I'm not good with Arduino).

Magnus

im also using MyFocuserPro2 and have noticed this same issue has returned..

i found by adding a blacklash of 1 in the focuser driver it seems to work ok now, see scrnshot.

just noticed one more thing, when i run autofocus all works well, but if autofocus runs through through scheduler the issue re-appears....i set "AF Overscan" to 0 and now works under both conditions.