Recurring focus issues - some days all is well and others not so much. But curious why SGP can report 98% percent success and then return a Failed focusing error… too many range shifts.
Thanks for any input.
From the screen shot:
- Your step sizes are too big. I would try with them cut in half
- You have backlash (“dog ear” on right hand side of the graph). I’d increase the amount of backlash compensation you are using (or turn it on if you haven’t already done so)
I agree with the suggestions from @SteveWinston above.
The quality factor was high after 6/7 points (= 85% complete ), presumably it dropped below the quality threshold after all 7 points, which is what counts. It would be clearer if you could send us the log file with these runs so we can runt it through the SGP AF Logviewer.
While the answer has been given (adding backlash of at least one step size - the focuser is just eating up the backlash in the first step so no change in HFR value), it is a little frustrating when one sees a graph that intuitively seems pretty reasonable (i.e it seems very clear where the focus point is) only for the focus algorithm to fail and go again without testing a focus point or rejecting outliers that might arise from backlash. It’s not that the focus routine doesn’t work (it works well when all is sorted by the user) it just seems a little unintuitive. I’m not a coder of any sort so maybe this is massively tricky but have found myself yelling at SGP to just ‘see’ a focus point that seems very obvious to me (as in the graph in the original post).
We’re not seeing the full or final graph that caused focus to fail in the original post - we would need the full logs to see that.
It is likely that the final point was sufficiently off the curve that SGP no longer felt confident that sufficient quality had been reached.
I think the key thing is though that once you dial it in it is consistent, at least for my setup. I have it running multiple focus routines per night across multiple nights, and I do not get focuser failures.
1 Like
Yep. I’m happy now mine is set (no failures) but did go through a process to get there (the yelling bit and then calmly adjusting backlash!). It’s not that anything is wrong with the AF routine but there’s a point where intuitively you can see the result but the algorithm can’t. This, of course, may be unique to my setup and seeing conditions, but the original post reminded me of what I was seeing and feeling about it.
Obviously I don’t know your particular hardware setup but I was getting exactly the same issues - running a ZWO EAF - days (nights) of frustration. In the end the problem was not SGP at all but the grub screw that tightened the adaptor onto the telescope focuser shaft was very!! slightly loose and allowed the shaft to move ever so slightly in the connector - particularly when the focuser changed direction. Tightened the grub screw really tight and everything started working immediately.
1 Like
Yes, seen that with my Lakeside focuser too !