The linear algorithm was developed to overcome certain issues that people experience. More likely for when there is a certain amount of backlash and perhaps when you're not using a stepper motor. It takes longer to achieve focus, but I think people get better results.

If your AF has a stepper motor, and the backlash is minimal, or able to be compensated for, then I would normally use the Polynomial algorithm (which I use now, and usually get good results).

I have DeepSkyDad AF3 focuser with NEMA stepper motor. I will try to test both algorithms tomorrow

Hi All,

I changed over to running EKOS around a year ago and in that time I have not been able to run the focus module successfully with my 8” RC, standard crayford focuser and iOptron EF. I have reached the point where I have good steps in and out, use linear algorithm and get a ‘sort of’ V-curve. All setting seem to be about right for my setup and visually Autofocus seems to drive nicely out-of and back through focus.

My problem is that the module fails with ‘fail to detect a star’ when the image shows an annulus - this is even stepping outside a nominal focus by orders of 10s of steps. The annulus is bright and tracked in the green reticule, but EKOS seems to not track the outer diameter and maybe is attempting to measure part of the annulus rather than the whole.

I’m any case, it will only focus around 1 in 20 or more attempts and even then tweaking does better.

After browsing many discussions on focusing and probably missing the important one, I’m stumped! Can anyone point me in the right direction to track the outer diameter of an annulus for successful focusing?

Many thanks

Mark

You have to adjust the Sep star settings to allow it to recognize larger star sizes. But even then, you shouldn’t have out if focus star sizes larger than 2 or 3 times the HFR size of an in focus star. Do if your in focus size is 1.5 then your largest out if focus sizes should be no more than 3 or 4”. I get great fox’s results thus way. I never let it focus so far out as to show a star as a donut.

Thanks Peter,

That's useful information. I should note that it's EKOS that pushes the focus to show donuts - I focus to a good level (just a smidge focus in from ideal focus) before I autofocus and even reducing the step size to between 10 and 20 steps for the iOption stepper motor, with a multiplier of 1 or 2, EKOS stills steps away from focus sufficiently to produce donuts on my RC. I'll try the Sep setting when we next have a clear night.

Many thanks

Mark

Then your step size is too large. There is no one-size fits all, as it all depends on your focal length, motor resolution and focuser. But you need to know how far the focuser must move to increase the HFR by 2x the perfect focus HFR. Then you can divide that by the step size and multiplier to move the focuser out by that amount. I can sho you my settings, but they will not be the same needed for your setup. But once you nail down that setting, EKOS does very well at focusing. I've used the Esatto, ZWO EAF and Moonlight focuser plus the moonlight Litecrawler. All work great once you get it to swing 2X the HFR.





PS: I am using a C11 edgeHD at f/7 (1960mm focal length) with Moonlight Litecrawler and ASI1600MM Pro,)

Thanks Peter, too big a step size has been my playground for a while, but I’m only using 10 and 15 steps at present. I can see the star image grow and shrink nicely, but it’s EKOS that does not seem to stop before it goes so far as to produce annuli. I don’t get this at all and it’s going past a x3 HRF.

I will keep playing, but one thought that is coming to mind is that the iOptron focuser is not up to the job…

Cheers

Mark

Try setting the max travel to restrict how far it moves from its initial start point.

I’ve been trying to find how to handle donuts with my C6 when it moves even a 100 steps from its focus point and this seems to at least stop the algorithm wandering off. Unfortunately not sure, at least in my case, whether this will stop me getting a V to identify the new focus point.

As Nigel says, your “max travel” is set too high.

Hi Peter

I was just looking at your settings and thinking the sizes were huge, then I checked the lightcrawler steps per revolution - 374920 so rather larger than my Celestron focus motor which has 1000.

I thought my max travel of around 100 was small (nb still waiting for a clear night to try my settings out properly) at 36 degrees of rotation but yours seems to be around 2 degrees. Am I missing something because that would seem to mean it’s really hard to manually focus as just a tiny twitch would be 2 degrees?

Regards

Nigel

Thanks Nigel and Peter, max travel is probably the culprit as you say, but based on experience of other autofocus programmes, I was expecting EKOS to determine a limit of increasing size itself to decide whether to focus in or out, to auto v-curve the action.

Is there a description of the algorithms employed in the various measurement options so I can see how the main variables the user inputs will affect the operation of the system?

All the best

Mark

Yes, the Litecrawler has extremely fine steps, for long focal length scopes. While I don’t manually focus at all, it has a dial to adjust the step size increment so you can move the focuser quickly if needed.