Autofocus miscalculation

[Orfeas Voulgaris]

I use myFocuserPro2 autofocuser on a SW Explorer 750/150 PDS. After updating from v1.10 HF1 to HF2 I noticed autofocus doesn’t work as it should. With good seeing conditions and a cloud-free night it would ruin focus position. I attached the full log. The problem occurred 2 times, at 23:02:45 and at 01:33:22. I manually started the autofocus procedure at 23:18:59 and it failed again, unable to get a valid V curve. After checking tension on the focus knob and screw tightening (drawtube slipped a little as you can tell from the position shift), I ran it again at 23:25:10 with a successful outcome and things looked ok but when it ran automatically again at 01:33:22 it totally ruined every shot afterwards with HFR jumping from 2.46 to 11.1!!

This is an excerpt of the log file, the first time this happened:

Autofocus @ sequence beginning:

[2021-04-11T21:30:15.0945]   [DEBUG]     [Message] Current Focus: Position: 7316, HRF: 1.56122118587333

Last file saved before autofocus:

[2021-04-11T23:02:28.1421]   [INFO]      [Message] Saving image at C:\APT_Images\Cigar Galaxy(ha)\2021-04-11\LIGHT_0018_2021-04-11_22-57-24_300.00s_2.58hfr.fits

Autofocus start:

[2021-04-11T23:02:45.1591]   [TRACE]     [MemberName] StartAutoFocus

Autofocus ends with this message:

[2021-04-11T23:04:46.1920]   [WARNING]   [MemberName] ValidateCalculatedFocusPosition
[2021-04-11T23:04:46.1920]   [WARNING]   [FileName] E:\Projects\nina\NINA\ViewModel\AutoFocus\AutoFocusVM.cs
[2021-04-11T23:04:46.1920]   [WARNING]   [Message] New focus point HFR 1.7834269278545 is significantly worse than original HFR 1.48403394374643

[2021-04-11T23:04:46.1920]   [WARNING]   [MemberName] StartAutoFocus
[2021-04-11T23:04:46.1920]   [WARNING]   [FileName] E:\Projects\nina\NINA\ViewModel\AutoFocus\AutoFocusVM.cs
[2021-04-11T23:04:46.1920]   [WARNING]   [Message] Potentially bad auto-focus. Restoring original focus position.

[2021-04-11T23:04:46.2490]   [INFO]      [MemberName] MoveFocuser
[2021-04-11T23:04:46.2490]   [INFO]      [FileName] E:\Projects\nina\NINA\ViewModel\Equipment\Focuser\FocuserVM.cs
[2021-04-11T23:04:46.2490]   [INFO]      [Message] Moving Focuser to position 7316

Image right after (failed) autofocus is this:

[2021-04-11T23:09:54.0036]   [INFO]      [Message] Saving image at C:\APT_Images\Cigar Galaxy(ha)\2021-04-11\LIGHT_0019_2021-04-11_23-04-49_300.00s_4.60hfr.fits

‌

‌

[Stefan B]

Autofocus writes separate json log files. %localappdata%\nina\autofocus

Please include your autofocus runs from there to quickly see the problem.

[Orfeas Voulgaris]

autofocuser json files

[Stefan B]

After looking at the curves there are some artifacts inside, which could indicate potential backlash.

What is your backlash compensation method set to in nina and what are the values for IN and OUT? Additionally is there any backlash compensation set inside your focuser driver?

[Orfeas Voulgaris]

The backlash setting is 20 steps inside the driver. Nina compensation is set to 0. However the exact same setup was functional on 8 & 9th of April at home and 10th of April after the update at a remote (suburban) location. I will post those logs tomorrow. If nothing else seems suspicious software-wise I will check again if all connectors make good contact.

‌

Thank you for your immediate response!

[Orfeas Voulgaris]

The logs from 8, 9 and 10th of April each day in a separate folder.

[Stefan B]

I’m not too sure why the curves are looking so differently. You need to check in detail on your settings if stars are properly detected and that there is no slippage as well as no backlash anymore in the focuser system. For further help I suggest to hop onto the discord chat, as there is a dedicated channel just for autofocus discussions.