INDI Library v1.9.4 Released (17 Jan 2022)

Bimonthly Stable INDI Library release introduces new drivers and fixes for existing ones. Some highlights:

Auto focus Newbie. Need help understanding INDI control panel for ASI EAF

  • Posts: 6
  • Thank you received: 0
Really excited to get an EAF the other day.  I received the ZWO 5v EAF and attached it to my Radian Raptor. It's cloudy so I can't really test it out like I want, but I am already stumped by the control panel and the "relative position", "max position", "absolute position" etc.  Trying to find literature on it but only found the focus module instructions for EKOS. Can anyone explain or point me in the right direction to some reference material that will explain this better. I'd like to get the settings right so I don't damage the motor in the EAF. Trial and error makes me nervous with sensitive equipment like this.

5 months 3 weeks ago #74078

Please Log in or Create an account to join the conversation.

  • Posts: 406
  • Thank you received: 150
The control panel might look daunting, but it's quite simple once you understand that there is only one thing the focuser does, moves in and out :) All the controls (apart from temperature reading if you have the sensor) are related to that.

The focuser position is defined by steps which are the numbers you see, zero usually means that the focuser is fully retracted and goes out when the step count increases. EAF is an absolute focuser so it always moves to an absolute step count / position, the relative position just adds/subtracts from that count so if you are at for example 5000 and move 100 out, the focuser moves to 5100 and so on.

Maximum position defines how many steps it can move out before the focusing range ends. This is mostly just a safety feature so it's not critical if it's too large unless you try to move out too much and the focuser stalls and EAF loses steps and goes out of sync. The step range depends on the focuser (not EAF) so you need to test that out with your own. Some focusers have a homing function that finds out the limits, but EAF does not.

The initial steps for you would be something like:
- move the focuser manually to the fully retracted position before connecting the EAF for the first time. EAF does remember the last position it was in so if you don't move the focuser between uses, even if you turn off power, it's always ready, but otherwise it's easiest to start from the zero position.
- connect EAF and if field Absolute Position shows something else than zero when the focuser is fully retracted, use the sync control to set the count to zero
- try to set the absolute position to something like 500 and see if the focuser moves out or not. If it doesn't, set the position back to zero and enable Reverse Motion in the control panel and try again. Some focusers like mine require this, some don't.
- at this point you should be able to move the focuser back and forth with repeatability

To find the maximum step count you'd just increase absolute position until the focuser doesn't move out more and that would be the maximum position you want to set (or a bit less for good measure).

Backlash is the amount of steps it takes for all the gearing in EAF and your focuser to re-engage after changing direction. It helps repeatability in auto focusing (unless you use the linear focusing algorithm which compensates for this itself) so it's worth the time to tune the value. It's mostly a trial and error thing, but you can get close enough value by moving the focuser by small increasing step counts back and forth and find out what is the minimum step count that actually physically moves the focuser axis and that's your backlash value. With my TS 2.5" rack & pinion focuser it's around 90 steps.

For actual auto focus in Ekos there is a lot of discussion on this forum, but that's also something you need to tune to your particular system. The default values are probably a good starting point, but important values are those dealing with step counts as those are focuser specific. Also you need to have decent initial focus so that the algorithm can find the stars and detect differences in their sizes. Also exposure time and gain should be such that the star doesn't get clipped but also isn't too noisy or too affected by seeing. I usually use a couple of seconds for LRGB and a bit more for narrow band filters with my mono camera. OSC cameras work a bit differently and I've found it helps to use 2x2 binning for focusing with them.

The initial step size should be such that from fairly good focus you get to somewhat bad (but not too bad so you lose the stars or they turn to donuts if you have a reflector scope) focus so that the stars grow enough for a good V-curve. Max step size and max travel are safety values you should set high enough for the focus to work but low enough to abort if the algorithm goes haywire. Here my own values for initial step size is 200 and max travel I've set to 500 as my focus is never that much off that moving 500 steps to any direction would make it better.

Also if you want to try the linear focus algorithm, the initial step size value might need to be larger than for the polynomial algorithm. I find polynomial works faster for me as I have all the parameters tuned over the years, but linear algorithm might be easier to get working at first and it also handles backlash automatically as it always focuses inwards.

The rest is for you to explore once you get your gear under a clear sky, good luck :)
Remote observatory running Ubuntu 21.10 x64 with ScopeDome 2M, Skywatcher EQ8, TS 8" RC, Atik 383L+, ASI178MC, ASI120MM-S, SX Lodestar X2, SX USB wheel with OAG, ZWO EAF, Gemini SnapCap, USB_DewPoint, KomaHub
The following user(s) said Thank You: jiberjaber, Miguel , Rafael Schlegel, Justin
5 months 3 weeks ago #74149

Please Log in or Create an account to join the conversation.

  • Posts: 6
  • Thank you received: 0
Really helpful. I appreciate you taking the time to explain all that. I'm sure I would have searched about 100 sites to put all that info together so nicely. I can't wait to try it all out, and let you know how it went.

Many Thanks
5 months 2 weeks ago #74153

Please Log in or Create an account to join the conversation.

Moderators: Radek Kaczorek
Time to create page: 0.826 seconds