TheSkySearchers

have no idea where to put this post, we have a guiding sub forum but no auto focus lol

for some reason sgp autofocus is taking its time ,it will do 7 points in and 7 points out,it's start position will deviate by 150 -200 from the filters focus point then work its way back in into focus then out again to the 150-200 + position,

the + 150/200 is basically in no mans land with large donut stars ,i don't really see the point for starting way out there, usually 50-80 covers the temperature over the nights. so is there anyway to change the starting points for the auto focus ? is it in correlation the the min and max step size ?

my filter list are pretty good it's just auto focus wants to take a large step to find the out of range area...

Follow the autofocus setup in the AGP manual: https://www.mainsequencesoftware.com/Co ... Focus.html. It tells you how to determine the proper step size for your equipment.

Don't forget to set the backlash, too. I have found that a healthy backlash setting is slightly larger than your step size. You can detect backlash if the V-shaped focus graph has "wings", especially on the right side.

I started with the advice in Kathy's link and adjusted it from there. I do a total of 9 steps. 4 bottom 4. Works well for me.

But whatever number of steps you choose make sure it is an odd number.

Also make sue that your exposure duration is long enough to get a good star field. I use 12 seconds.

KathyNS wrote: ↑Wed Dec 04, 2019 5:31 pm Follow the autofocus setup in the AGP manual: https://www.mainsequencesoftware.com/Co ... Focus.html. It tells you how to determine the proper step size for your equipment.

Don't forget to set the backlash, too. I have found that a healthy backlash setting is slightly larger than your step size. You can detect backlash if the V-shaped focus graph has "wings", especially on the right side.

Kathy, what do you mean by "wings"?

Cheers, I'll look into the manual, Kathy means wings as in with the data graph will pan side ways flat at the start and not have a linear progress with relative of fwhm and focus position, there's a flat deviation because of backlash.

UlteriorModem wrote: ↑Wed Dec 04, 2019 5:38 pm I started with the advice in Kathy's link and adjusted it from there. I do a total of 9 steps. 4 bottom 4. Works well for me.

But whatever number of steps you choose make sure it is an odd number.

Also make sue that your exposure duration is long enough to get a good star field. I use 12 seconds.

Thanks for the pointers. Wow 12 Seconds! I use 2-3 seconds exposures

Narrow band filter

6 seconds is probably fine with a lum or no filter.

What f ratio are you at?

Stuart wrote: ↑Wed Dec 04, 2019 6:13 pm Kathy, what do you mean by "wings"?

The rightmost point on the graph is out of line with all the others. If it was consistent, the HFR value at the position should be between 7 and 8. The fact that it isn't is due to backlash.

KathyNS wrote: ↑Wed Dec 04, 2019 8:34 pm

Stuart wrote: ↑Wed Dec 04, 2019 6:13 pm Kathy, what do you mean by "wings"?

Untitled.png

The rightmost point on the graph is out of line with all the others. If it was consistent, the HFR value at the position should be between 7 and 8. The fact that it isn't is due to backlash.

I see. And what backlash setting do you use? Is it "in" or "out"? I've never done this before.

Stuart wrote: ↑Wed Dec 04, 2019 9:18 pm I see. And what backlash setting do you use? Is it "in" or "out"? I've never done this before.

My best step size worked to be 90 units. Because my results were indicating that the backlash was almost the size of a full step, I set my backlash compensation to 100.

The direction depends arbitrarily on your own specific equipment. I know that my temperature compensation ends up setting the focuser to lower position numbers as the temperature drops. Since it is lengthening the optical path to compensate for contraction of the OTA, lower position numbers therefore mean an outward motion. Therefore, the initial starting point for a focus run, which always starts at higher numbers, is inward. Your focuser may operate in this direction, or it may be reversed. You will have to determine that for yourself.

Knowing that it starts with an inward motion that is not far enough in due to backlash, I need it to go farther in and then come back out, so that the first data point is measured with the focuser moving in the same direction as for the rest of the focus run. So I have the backlash compensation direction set to OUT.

If you determine that higher position numbers correspond to inward motion, you would need to reverse this setting.

Thanks Kathy, got it. You have officially made it into a "tutorial" bookmark! :-)