Autofocus - first point out-of-line with others


#1

I’ve made good progress with autofocus - I was expecting it to be difficult given that it is a C11 with Starizona Feathertouch on main focus knob. However, often the first (and sometimes second) point is not in line with the other points on that side of the V. Looking for input from other users with similar equipment for ideas on how to address. Or perhaps this is as good as I should expect?

Here are my settings:

Thanks in advance for any suggestions.

Dave NL


Determining the value to put in for focuser backlash compensation
#2

That looks like backlash to me. Are you sure 250 steps is adequate? Maybe try something like 500 and see if that fixes it.

Thanks
Jared


#3

Thanks Jared, I’ll try that out next clear night.

DaveNL


#4

Another possibility is that SGP is picking up a hot pixel, thinking it’s a star. When this happens, it has the most profound effect on composite HFR when focus is farthest from center. You might try increasing your Minimum star size.

Tim


#5

Oh, and did anyone else try to drag the warning box away to see what the bottom of the focus curve looked like?

:slight_smile:

Tim


#6

This is pretty unlikely for a couple of reasons:

  1. we reject single pixel “stars”
  2. A single point among 300 won’t change the average that much

Thanks,
Jared


#7

First I would make sure the main focuser knob on the sct is turning counter clockwise during the focus curve - so it is pushing up against gravity.

Second - the entire focus curve appears to take about 1000 steps. Each person’s step size could be totally different from another’s - but the curve itself normally requires just a small turn of the main focuser knob. So if you only have 500 steps of backlash, it isn’t removing much backlash at all. I would aim for 1/4 to 1/2 turn of the focuser knob to remove backlash - which for your system may be about 2000 steps.

A lot of this stuff you tell by ‘feel’ if you remove the autofocus control and just turn the knob clockwise and ccw. With the scope aimed up you can turn it clockwise for a bit - and then start turning it ccw. You will feel it take some time to fully engage - and that gives an idea for how much backlash you should remove.

Frank


#8

Thanks to everyone for the suggestions. Based on past experiences with this motorized focuser, live-view DSLR and bahtinov mask, my “feel” for the backlash would probably be about 200, but perhaps I am underestimating. Last night I tried increasing the backlash compensation to 500 and things seem to improve. Some were near perfect:

while others were still not quite so:

Next time out (hopefully tonight) I’ll try something larger like 750 or 1000.

Some additional questions:

  1. Sometimes, the last point on the left is “out of alignment” (middle graph above is a good example). Is this also backlash or possibly something else?
  2. Is there an easy way to identify which way (ie. In or Out) is counter-clockwise? I’m assuming I’ve got it right or my graphs would be much messier than they are now? I guess this is something easy enough to test …

DaveNL

PS. Overall, I’m pretty impressed with this autofocus thing.


#9

Just watch which way the focus knob is turning as it goes through the autofocus steps. It is best if it is counterclockwise.

Frank


#10

With the Starizona Microtouch autofocuser, I can’t see the focus knob. Actually, it replaces the focus knob - and the encasement does not provide visibility to any moving parts :=(

DaveNL


#11

I have an Edge 9.25 with the Feather/Micro Touch too. My compensation step size is 300 and the compensation direction is IN.

Dean


www.mainsequencesoftware.com