Focusing problems

Description

I switch the focal length of my telescope with a reducer/flattener and can’t not get the focus function to work properly in fitting a curve. I have put all of the information in the video I made. I used a small step size and a very large one and I still got errors. I got excellent focus and images but I did my focusing in @ focus 3. I would like to keep everything in SPG. Can anyone help point me as to where I went wrong? Thank you.

Video of the focus run

Useful Info

OS:
Ver:
I have version 3.0 SGP

The video is over very quick and you do not show many data points. However I would say that you are in the right region with step-size of 250 but I think that run has most likely failed due to uncorrected backlash at the first data point. I would say your backlash setting needs to be increased to circa 300.

I would recommend that when you think you are at good focus (HFR c1.0 - 1.2) that you then:
a) Use The Focus Control/GoTo function to move the focuser out 750 steps and then check that the HFR measurement has increased to circa 4.0.
b) If HFR is not around 4 then move the focuser in/out as appropriate until the HFR delta is about right.
c) Assuming that the 750 move is correct then change your AF settings to step-size 250 with 7 steps. This will give a few more data points for you to verify that your AF curve is showing a reasonable parabola shape.
d) If the first data point is still hanging out to the right, increase your backlash setting some more.

Once you have established the approximate distance that you must move the focuser in/out to get a reasonable HFR delta you can refine the steps/step-size settings as you think appropriate.

Ok thank you for letting me know to check the backlash. With my moonlight on my other telescope I had likely turned it off. I had calculated the focus zone with the below formula. I don’t understand where I went wrong with the number I calculated and why I am having to use such a large number.

I appreciate your time.

Joe

Look toward the bottom for critical focus zone.

Hi, I may be missing your reasoning here but I understand CFZ to be a small region either side of ‘perfect’ focus within which the human eye is unlikely to see any change in focus accuracy. However when using SGP AF you do need to move the focuser out of focus by a decent amount in order to get a clear change in HFR values for plotting the focus curve. I think in your first run with step size 25, the focus range was too small to give the AF routine a clearly detectable inflexion point at best focus. The second run looked better but as mentioned above I think the curve was poorly defined owing to backlash. You might have got a decent AF result with 7 data points but 5 was insufficient. For what its worth the SGP help file says to look for a change in HFR or 3-5 times the lowest achievable HFR value. My experience was that with a 5 times change (using L filter) I was very often getting very few star detections at the extremities when using RGB for a reliable results so eventually went for a maximum delta of 3-4 times the minimum HFR. At best focus with my set-up I was getting best focus HFRs of 0.8 - 0.9 in good seeing and so chose a step-size that took me to c3.6 - 4.0 at the extremities of the curve.

Hope you soon get some decent results.

Thank you I will see what I can do tonight. I am going to add to my Rosette. I would like to keep everything in SGP without having to fall back to TheSky. I think the backlash change will help. Took the enclosed last Thursday. 71 subs of 2 minutes each with the L Enhance filter.

Will let you know how it goes.

I increased the backlash as you advised and experiment with the step size. I went with 275.
Everything worked well last night. Thank you.

Hi Joe,

First of all, congrats on your Rosette image. You have some excellent detail and I like the nebula colours.

Looking at your AF curve, the first data point on the RHS looks ‘limp’ and the overall AF curve is not fully symetrical. This is appearance is indicative that your focuser backlash is not fully ‘dialed-out’. It seems to be fully gone by the second data point. As your step size is 275, I would add a further 275 steps to your current Focuser Backlash Compensation: Compensation step size. When this is done I think that your Focus Quality result will likely improve to c97-98%. On the graph shown I would expect the final focus position would move inwards slightly. Would this change the validation frame HFR? Maybe hardly at all as you look already to be pretty much in the CFZ.

Clear skies

Mike

1 Like