Hi, I got the Ekos focuser working very well on my refractor but (a) that was a while ago, (b) I upgraded StellarMate and didn't manage to transfer all my settings, and (c) this is a new telescope and focuser setup anyway.

I'm running an AT8RC with a DIY direct-drive motor (PG27). The RC is 1625mm focal length and f/8, so the CFZ works out to about 65 microns. One step on the motor is about 2.2 microns.

From the days when I was trying to build a myFocuserPro, I recall the advice that your motor's step size should be no more than 1/10 of the CFZ if possible. But the step size the focus routine uses is something else again. Messing around pushing the buttons last night, it seemed that to be readily visible in HFR values (which were quite noisy), I needed to use a step size on the order of the entire CFZ!

So an initial step size of maybe 30, focus multiple 5? I'm using SEP detection, 1-Focus-Default profile, Linear 1-Pass, and a hyperbola curve fit per the docs (which are well in advance of the Focus module in StellarMate, but that's perfectly fine). I also poked in an AF Overscan of around 30, since I remember that's about what the old rig had for its backlash param in the driver (same motor, though pulley instead of direct-drive and rack & pinion instead of Crayford).

I suppose I could average one or more frames to get less jumpy HFR values. Would make this manual phase even more tedious, but once the AF is working it would be less of an issue.

Any suggestions gratefully accepted.

You seem to have gone into it much deeper than I did when setting up my autofocus, I read about the CFZ but never worked it out for my particular scope as I managed to get everything working quite reliably.

The way I have been setting it up is to get the position of perfect focus with a bahtinov mask and taking a note of the absolute step number, use this as a starting point when trying to find the right AF routine settings.
After that it was a matter of trial and error to get the following results below during the autofocus routine.
1. If the HFR is 1.0 when in perfect focus you want the initial sample HFR while doing the routine to be 4x-6x this number and let it work its way back and make the curve.
2. At the very bottom of the curve you want 2 sample points either side, if you have 3 at the very bottom your step size is too small (1st pass only).

So by adjusting my initial step and step size I could get a nice curve using this method.
It takes me between 14-25 iterations with 1s luminance, F7 refractor, 2600mm camera pointing pretty much anywhere in the sky, from there I have my offsets for the filters and its usually very stress free.

I tried to do the multiple exposures per step but it wasnt very good in my opinion, took a long time.
I also have set 0.5s settle time between adjustments.

I've used this method on both my SCT and Refractor.

Happy to get a screenshot of my settings if your interested to have a look.

Hi Rick,

Let me give a little plug to Focus Advisor...


It will suggest basic params to get you setup. You will be able to improve things with some tuning later but should get you going. Select the Advisor tab in Focus and hover the mouse over each label to see the suggested settings. Check Update for the groups to be updated and press Update Params.

There are a couple of settings you will have to work out yourself like Initial Step Size (setting equal to CFZ should be a good place to start), Camera gain and Backlash.

If it works for you that's great, if not let me know and I'll tweak it.

The settings you mention in your post are sensible IMHO, so good luck!

Thanks, John. Onward!

Hy: I can try upgrading KStars to get the Advisor, but I'm not sure it's in the StellarMate repos yet. Reading the docs I did immediately go "oooh, I need that!". Looks like just a great tool.