Hi John,
I think it is a good thing that the hard stop position is 1500 motor steps from the focal point. That allows you a lot of plenty of room outside of focus for AF to work the routine. When I first set up mine, I was at about 1000 steps from focus to the hard stop. That has changed as I zero in on optimal backfocus position of my camera. With the light pollution/UV/IR cut that has moved my camera out a bit and now I am down to an optimal focus position in the range of 300 from the dead stop of 0. And that is with my stop moved to allow extra rotation outside of focus. But I know that I am close. Only a little tilt to resolve, if I choose to deal with it.
Yes the motor is powerful. I saw mine actually flex the plastic mount to the Astrodymium mount rings, almost appearing to flex enough to break the motor mount, but with the angle generated, the belt just slipped the cogs. And this was when I was using the manual motor control. So stupid on my part. Had this not occurred, I am sure something would have broken, possibly the motor!
It is odd that you had the missregistration at the end of your AF routine. At 1500 ticks from focus point to camera focus stop, there should be no way you ever impacted the camera stop. I am not sure what you mean by:
John Noble:
What I’ve read says the initial move is 150 steps then it comes back in 30 step increments then the fine focus goes out 45 steps and and comes in in 15 step increments.
And you do not say what software you use to do AF. I assume that you are reading the log. The first part makes sense, of what you say about moving 150 steps, then comes back in 30 steps, if you set your focus increments at 30 steps and that you have AF taking 5 focus steps to do one side of the graph (for example 5 steps to draw the graph for the inside of focus point). The second part seems like you have your AF set up wrong. Once the AF goes to a position inside of focus and if you have your AF and motor set up correctly, you should have it overshoot to eliminate backlash (have you done this?). [Setting up this overshoot backlash correction should/can only be done in one direction. Never set backlash compensation both in and out. One or the other! For your Samyang, which has a hard focus stop, you really need to set this backlash step to the inside focus direction.] So lets say for your setup you have your rig at 1500 (which is close to ideal focus point) + 150, so the start position will be 1650 (inside of focus). 1650 will be the position where AF will take its first data point. When it first moves the motor to 1650,
if you set the backlash compensation correctly, you will see the motor overshoot 1650 by several hundred (a number you choose) and then move back to 1650. From then on, it will
only move in one direction during the whole AF routine (Unless the routine decides that it has to then move back to a higher position to complete a decent curve). Getting back to the beginning of the AF routine, then the motor will move downward in 30 tick increments, taking an image and hfr at each step. It will take 5 readings to get to 1500, plotting each point. It will then continue moving (going outside of focus) until it reaches 5 positions (150) outside of focus, or 1500-150 = 1350, plotting results each time. You should
never have the AF routine do one part of the routine moving in one direction, then finish the other part in the other direction. All these motors have backlash and you will never get a good or consistent focal point. Because of this, your statement that "then the fine focus goes out 45 steps and comes in in 15 step increments." makes no sense to me and suggests you have it set up wrong. Also, you do not need to do a "fine" focus, or corellary to that statement, a "rough' focus, then a "fine" focus. If you know what the focal point is, then all AF actions should be fine focusing. You start the AF routine knowing that you are close to actual focus. If it is not the last focus point you used during your last session, then you can manually get into the ball park prior to doing the AF. That you check during session set up. I would suggest that you go to Patriot Astro Youtube: Setup NINA Auto-Focus Quickly and Correctly. It is specific for NINA but has some good general points. With my 135, I use 40 tick steps. That takes my AF routine from hfr of ~5 down to below 2. If you think that smaller steps will yield "finer" results, that is likely wrong. Seeing, and other issues may make the sampling at smaller increments may just make it more difficult for the routine to get a smooth curve. I use 2 second exposures. The routine does not need to see every star and at f2 you will not be short enough stars for focussing purposes. I also only do 4 increments. Somehow NINA alway finds a way to insert more, so I don't give it any reason to spend extra wasted time. However, I also do two exposure at each test point. This comes from my other rigs that are wind sensitive, so I needed that to get better error bars for each point. But still the goal is to get proper focus without consuming a large amount of time. And about accuracy, AF algorithms will use mathmatical interpolation to find the focus from the graphical data. You do not need to choose points close together to have one of the points "hit" the exact focus point.
Oh, and I set my AF routine to only sample the inner 80% of the frame so that it does not have to deal with the the typically defective outer stars.
John Noble:
just curious if the HFR could yield even tighter focus!
I know that this was directed at SemiPro, but for my setup, this number is really not comparable between our setups, even though I am using the same lens and camera pixel pitch. Reason being is that hfr readings are sensitive to the camera gain/offset/debayer/exposure time and even the stretch used in the imaging pane of the software. However the best I got when I had it nailed the other night was 2.17. Then I rotated the camera and it went up to 2.6. I have been doing all this work at f2. If I cannot get this to work well at f2 when using my c-sized sensor, I can be sure that I will have no hope of moving to a full-frame camera. Currently I have hope! In any case, if you can get good hfr readings manually, it seems perfectly reasonable to do the focus manually. With my old cold hands, I spend way too much time fiddling with all that and never get it to be consistent. And with AF, I can do it from within my office at home! I used to like being outside with my equipment, but doing astroimaging outside no way compares to doing visual astronomy outside. For all this stuff, you still have to be looking at the damn computer screen and that blows your night vision. I can't even see the stars hardly under such circumstances so being outside is only nice when it is actually nice out! Honestly, when I spend my time outdoors during an imaging session, the best part is when the imaging session is set up, I can turn off the damn computer, and sit in my recliner with my binoculars, or just naked eye the sky. Motors, computers, rotators, etc. have long lost their luster to me for entertainment purposes.
