How do I calculate the slop in my focuser?

[tooth_dr]

Doing a bit of reading and although I’ve seen that backlash in a focuser must be accounted for, I’ve not seen a method of actually doing this.

I plan to run filter offsets for my ED80 using a basic SW autofocuser and Hitec Astro DC focuser.  I’ll start the night focusing on eg red using a Bhatinov mask and use offsets from there on to get better focus for green and blue filters.

My other scope has a Moonlite focuser and I’m getting a non symmetrical V curve, I’ve seen described as a seagull maybe (it’s flat at the top on one side), I now understand this is the backlash being removed between first and second points.  Then after the routine is complete my focuser goes back to the optimal point, but it hasn’t compensated for the backlash and therefore is not perfectly in focus. This matters at F2.8, as I can actually see the slight difference.

 

Im getting a sore head thinking about this, but would I be right in saying this is one-off setup process.  I just can’t see how to do it  

 

[bottletopburly]

Apt has a focuser backlash measurement tool 

[Jkulin]

Hi Adam, I used to use a piece of paper held into the focuser body and on a table I would slowly back the focuser out until the paper fell to the floor, that was my back lash figure, I don't worry any more as the Chromo's are all par-focal

[tooth_dr]

6 hours ago, Jkulin said:

Hi Adam, I used to use a piece of paper held into the focuser body and on a table I would slowly back the focuser out until the paper fell to the floor, that was my back lash figure, I don't worry any more as the Chromo's are all par-focal

Nice improvisation 👍🏼 Unfortunately it’s not the filters but the scope in my case. 

[Jkulin]

4 hours ago, tooth_dr said:

Nice improvisation 👍🏼 Unfortunately it’s not the filters but the scope in my case. 

That's what I am saying Adam, to work out the backlash of the scope sandwich a piece of paper between the focuser tube and the body and very slowly ease the focuser out until it lets the paper drop, take the measurement and that is your backlash for the scope, then when you input your focuser backlash you then have the figure.

Don't worry what I said about the filters, that was just a passing comment.

[tooth_dr]

31 minutes ago, Jkulin said:

That's what I am saying Adam, to work out the backlash of the scope sandwich a piece of paper between the focuser tube and the body and very slowly ease the focuser out until it lets the paper drop, take the measurement and that is your backlash for the scope, then when you input your focuser backlash you then have the figure.

Don't worry what I said about the filters, that was just a passing comment.

I know

What I meant was, even if I had the most expensive filters eg chroma or astrodons, it's my scope and it glass that's not corrected, and therefore doesnt matter if the filters are parfocal ar not, I'll still have to deal with the backlash.  I'll have to do the measurements as you described.

 

[Laurin Dave]

Hi Adam..  I’d have thought that the backlash was about the same as the number of steps on the flat bit ..  Just make your backlash compensation bigger than that and try again...   if it’s still got a flat bit increase it again ..Etc..   having it a bit too big isn’t a problem ..  Also worth checkIng that all the connections between the motor and focuser are tight ..  mine work lose over time

Dave

[Davey-T]

According to the guy at Lakeside the number of backlash steps was pretty arbitrary as long as there was some, think I settled on twenty from memory.

This is with stepper motor don't know how DC motors measure " steps"

Dave

Edited May 21, 2020 by Davey-T

[tooth_dr]

13 minutes ago, Davey-T said:

According to the guy at Lakeside the number of backlash steps was pretty arbitrary as long as there was some, think I settled on twenty from memory.

This is with stepper motor don't know how DC motors measure " steps"

Dave

I have one stepper and one DC motor.  The stepper motor is on the epsilon so i was hoping to get the autofocus function working accurately, as the CFZ is small at F2.8.  The DC focuser is on the ED80, and really just want to use that for simple offsets, to get closer to focus when I change filters as current I dont focus at all between filters.  I guess if I dont get the backlash sorted in it then it's no big deal, as long as I take up the backlash and always offset the filters in the same direction ie inwards.

My head is telling me that unless you know how much the backlash is, then how can you reproduce the position following an autofocus routine?  I understand that the backlash compensation has to be larger than the backlash, so that you go past your point and take up the slack, but surely the number of steps of backlash needs to be exacting, or how would you ever get back to the position that the focuser determined?  Am I over thinking this?

 

Here is an example from my epsilon, a screenshot I saved in December.  In this instance the curve looks good, but it didnt get back to a good focus point so I manually intervened afterwards. I also have since sent my Hitec Astro MHP v3 back to HA, and they replaced the internals to upgrade the size of steps, from 12 to 2um. This was important, as I believe the CFZ of the epsilon is 10um.

 

Does SGPro need to know the exact amount of backlash?  If not how does it work around this during autofocusing.  In this example my focuser goes from 860 to 800, assuming from OUT to IN, and determines that position 828 is the optimal focus point.   At the end of the run the focuser then moves back out 28 steps to what is position 828.   But changing the focuser direction means there is some slack to take up first.  This uses X amount of steps of the inward offset of 28 steps, so the new actual position is 28-X relative to 800.  If X is known then the focuser could move the additional X steps and end up exactly where it should.

How does SGPro work if it doesnt know X?

 

 

 

 

 

[tooth_dr]

1 hour ago, Laurin Dave said:

Hi Adam..  I’d have thought that the backlash was about the same as the number of steps on the flat bit ..  Just make your backlash compensation bigger than that and try again...   if it’s still got a flat bit increase it again ..Etc..   having it a bit too big isn’t a problem ..  Also worth checkIng that all the connections between the motor and focuser are tight ..  mine work lose over time

Dave

Thanks Dave, it has been a while since I used the Epsilon, so I dont remember the step sizes, but it may be 10 steps (from above chart) - does that seem right?

[Davey-T]

Can't remember how it works, presumably its got a reference point to offset from ?

Dave

[tooth_dr]

3 minutes ago, Davey-T said:

Can't remember how it works, presumably its got a reference point to offset from ?

Dave

Right, so it moves past the point by the backlash compensation eg 10 steps, and then back those ten steps, so that we will be short of the optimal point by X.  It then takes an image, calculates the HFR, and based on the graph, it works out the number of steps need to get it to the optimal HFR?  Is that how it works?  I like to understand how stuff works so I can set it up and know what it's doing!!

[Laurin Dave]

Looks like you used a step size of 10 with the Epsilon..   it might be different with a different scope ...    RayD has a done a couple of his Simplified Astro videos on setting it up and explaining all this much better than I can...  its giving me a headache   

[tooth_dr]

7 minutes ago, Laurin Dave said:

Looks like you used a step size of 10 with the Epsilon..   it might be different with a different scope ...    RayD has a done a couple of his Simplified Astro videos on setting it up and explaining all this much better than I can...  its giving me a headache   

I worked out 10 using the instructions in SGPro - i think the method was getting focus with a mask first then moving out to get an HFR x4 bigger and working the step size out from that.  The Epsilon the only scope I need to get this working.  Thanks for the heads up on the videos, will look now.

Does SPro do a second adjustment to get to the optimal position, after the over shoot to take up the backlash, as the first position wouldnt be correct?

[tooth_dr]

45 minutes ago, Laurin Dave said:

RayD has a done a couple of his Simplified Astro videos on setting it up and explaining all this much better than I can...  its giving me a headache   

I have just watched the video on BL, thanks for that, explained everything.  My sticking point was trying to figure out the exact backlash.  This isnt necessary.

 

I have finally got my head around it.  I will not bore you with the explanation, but now that I've considered it with various sketches, etc, I understand how this works (he says)

 

Just for amusement here are my working out 😂

 

Edited May 21, 2020 by tooth_dr