I had been focusing well enough over the last few weeks thanks to the help received here. Last night, I was imaging NGC7000 before it transitted with no issues. At transit time, I did a manual Meridian flip, plate solved and set up a new sequence and the focus was way off. I initially couldn’t get a value below 12, so I tried to get focus manually (Optek Lynx focuser) and was able to get the numbers to around 6. After two hours of running a focusing run and trying to get closer manually, I packed it in. Funny thing - at various points, some of the stars had an HFR of under 4, but the focusing routine never reported any values on the graph below 6. And at tiems, one position, 45318, for example, showed as 6 on the chart and in the very next instance of the same Run, it reported as “12.” Any thought here?
As I recall, you are using a refractor for imaging, and one of the first things I’d be looking at is why the focus is shifting following a meridian flip. I understand why that might happen with certain reflector telescopes (when I first received my SharpStar 15028HNT, the mirror movement was so great that I would have to refocus every time it moved, but that was a defect which FLO sorted out for me. I do run an autofocus after a meridian flip, but I’ll probably stop doing that as it never budges), but a refractor? It suggests that there is some flexure or slipping in the focuser system that, of course, might translate into autofocus issues - if things are moving during a flip, the focuser could slip at other times. I think in the previous thread, backlash was discussed and I assume that you have that implemented at an appropriate figure, but if the focuser is slipping, you lose that 1-to-1 correspondence between stepper motor position and focuser position.
Also, what is the autofocus metric, is it “Half Flux Radius”, “FWHM” or “HFD - ASTAP” as they are all reporting different things. I assume that the display within SGPro shows Half Flux Radius, but if the autofocus routine is using either PinPoint or ASTAP to measure star size, it will report something entirely different, so there will be a mismatch between the graph and the preview window.
I had a problem a few nights ago with my RedCat which I hadn’t used for a while, where autofocus was struggling. In those circumstances, you can’t beat going back to a Bahtinow mask to get a scope properly focused. I am not quite sure exactly what was wrong, but I think I’d changed the step size which was too big (new PC and I’d obviously not set the profile up correctly), so the V-curve was anything but! Last night, the thing ran flawlessly from start to finish.
Is that a Crayford? If so, I would agree with the other poster that it is probably mechanical shifting. This can be pretty random seeming as it depends on where in the sky you are pointing in terms of how the imaging train weight affects the friction of the focuser.
I used the same control system for years on many different focusers but they were all rack and pinion and have never had issues with focus.
Sorry to say, my feeling is that any focuser that uses friction has no place in imaging. I know that is not what people want to hear, but it is based on 30 + years of imaging experience.
Thanks very much for the note, My focuser is an Optec FocusLynx DirectSync SW30. Not sure if it is a Crayfor, but it has worked well for the past 18 months, and I believe to is still fully functional. Optec :: #19788 - DirectSync SW30 Motor
Right now, I’m just trying to find the best method to get it back to a good HFR value so I can get back to work.
It is not so much whether it is functional but the design of the focuser. Crayfords typically use friction wheels to move the drawtube instead of a rack and pinion gear. That means that they can slip under some circumstances and those circumstances can seem random since they depend image train weight as well as on where you are pointed in the sky. So what worked on one object might not work on another. What worked in one part of the sky might not work in another and what worked the first part of the night might not work during the second part. It can be very frustrating.
This might or might not be your problem, of course, but it is a possibility that should be looked at.
Thanks, Terry. I should have been clearer in describing the scenario - I had done the flip and gotten good HFR values (under 2), but then somehow managed to bump the scope and throw things out of whack. So at this point, I’m looking for a good strategy to get back to good focus, as I believe all of the gear is still working fine. Maybe I should try increasing the step size so the system checks at a wider range of values - or lowering the step size and manually resetting to the original good-focus position value? Any thoughts on that? A mask will be a last resort 
Focus in SGP needs to be “close-ish” before autofocus can run well. You can loop exposures and manually move the focuser and pull up the Image History or pull up the Statistics window and just keep moving until you get “close”.
If your previous settings were working ok then you probably just need to get back to close focus and go from there.
Jared
Ah, but haven’t we all done things like that? How many times have I tweaked the mount alignment to get close to perfect polar alignment only to trip over the tripod leg while moving away from the thing? It’s a nuisance having to do all this stuff in the dark 
Something else to watch out for - if the motor connects to the reduction drive shaft of a focuser mechanism, there is a good chance it will slip as the reduction drive is not geared, but a friction device in most cases. The better motor systems have a gearbox attached and couple to the main spindle of the focus mechanism.
Thanks, Buzz. Not sure what sort of drive the Optec SW 30 Lynx has or how to determine that, nor how to trouble shoot it if needed. But the focuser has been working well enough for the past few sessions.
Reading from the manufacturer’s site “The motor assembly replaces the coarse focus knob of the focuser pinion block for Sky-Watcher refractor telescopes.” We can probably rule it out, but take care with couplers - the one on my Tak worked loose and I needed to find the ‘flat’ and tighten it up again.