10 replies to this topic

[physics23]

Hi everyone,

 

I have a Canon T2i DSLR and I use KStars/EKOS. With my telescope (SkyWatcher Evostar) I could set up autofocus alright. I would like to set up autofocus also for my Canon EF 200 mm lens. When connecting the camera, the dialog says that the Canon focuser is online, However, I don't seem to be able to move the focuser at all. I tried changing the step size, with no result. Did anyone have this issue/succeeded in setting up the Canon focuser?

 

Thank you for any help!

[t-ara-fan]

Try setting the switch on the lens to AF if it is on MF. My Canon 200mm f/2.8 needs it set like that.

[physics23]

Thanks, it is already set on AF

[jdelburgo]

Hi,

 

I had the same issue. I found the solution in the last post of this thread:

 

https://www.cloudyni...focus-function/

 

Best regards

 

José C

[physics23]

Hi Jose, thanks a lot, activating the viewfinder did work!

[m.takahasi]

I have no knowledge of the combination of EKOS and indilib.
Indigosky works by just selecting the focuser driver.
The cameras are 450D and 600D.
I had a phenomenon with the body and lens combination that the lens moves when the shutter is released.
I have set the "Back Button AF" in the camera options and the problem no longer occurs.
This was also the case when using the LensAF plugin in NINA

[physics23]

Just an update: now I can control the camera lens using EKOS, thanks to Jose's solution. But it looks like autofocus doesn't really work well. I set the step size such that the curve now is perfectly fitted by a parabola. But when moving outward to the solution, it doesn't get even close to the correct position, it overshoots by at least 4-5 steps. This happens for any value of overscan I have tried. I guess the focuser is maybe just not precise enough/has too much backlash. Did anyone have a similar issue and managed to solve it? Right now, I can just put on a Bahtinov mask and  focus manually, which is not too bad, but not ideal. Thanks!

[m.takahasi]

Is there any setting for overshooting method in Ekos?

NINA , APT and SharpCap in windows, Indigo-sky in Linux can be set up.

If we make full use of this, each point will always be accessed from the same direction.
You can cancel the backlash
Camera lenses also have a slight backlash, depending on the model, so setting up 2 or 3 steps, etc. will always give you the best AF.
I remember that Canon's ultrasonic motors were the only ones that had almost no backlash.

Edited by m.takahasi, 01 June 2025 - 05:22 PM.

[physics23]

Yes I did try many different settings for the overshoot (called Overscan in EKOS). But none seem to solve the problem. In fact, with zero overshoot, when focusing to the solution it overshoots by a lot. So I'm not sure what the issue is, this doesn't seem what would happen due to backlash? (this is an ultrasonic Canon motor)

[m.takahasi]

Do you have a Windows laptop?
If so, try using the live view screen with Canon EOS Utility.
Does the same misalignment occur with ground scenery?
If so, how many clicks are required at the smallest step size to correct the misalignment?
(There are three types of manual focus buttons on the screen: <<, <<<, and IN/OUT.)
I did not notice any backlash with the Canon 70-200mmL.
However, with the older Sigma ED 300mm lens, I needed to click back three times at the smallest step size.
That's why I'm using the overshoot method.

I don't know anything about EKOS, so maybe an expert can help you.

[FrostByte]

Which autofocus algorithm are you using? Most of them work best with an absolute focuser, but the autofocus motor in a lens has to be treated like a relative focuser, meaning it does not know its absolute position so cannot go to the minimum by itself. I don't have my mini PC with me so cannot open EKOS, but you may have better success with one of the one-sided algorithms depending on what you've been using already.