Hi Jared
Thank you for the clarification.
I’m wondering if it is better to apply backlash compensation only when the focus direction is changed during autofocus runs to reduce unnecessary compensations.
Thank you.
Probably not. Backlash exists generally because of outside factors like gravity or mirror flop. If it’s true gear backlash then it’s likely fairly small and the amount of extra steps needed to remove it is trivial as far as time is concerned. The main issue is with SCTs and mirror flop which can take a substantial amount of travel to remove. Thus you always want to end the move in the same direction. But that should always hold true no matter if it’s mirror flop or gear backlash.
Thanks,
Jared
…old topic, but still I dont get it.
I dont need backlash compensation from gravity, I need backlash compensation from gear backlash which is in my case 20-25 steps FOR EVERY direction change. It doesnt matter if its in or out, it happens on every direction change.
I guess SGPro is of no use to compensate that…
Look at freestar8n post below:
Continue reading that thread because Frank described well how to auto focus using stock focuser of SCTs.
Peter
…I apologize, still I dont get it, I have no SCT, I have a refractor with a 2.5“RP focuser…
I‘d really like to switch from APT to SGPro just because of the autofucus feature, but it seems like there is no way of gear backlash compensation.
The documentation/handbook also fails in explaining the backlash topic in detail.
APT has the gear backlash compensation but a bad autofocus method…
…makes me feel pain…
OK, I assumed this thread was about focusing for SCT scopes but it should not matter.
I also have refractor scope and my auto focusing compensation is set to “IN” meaning that anytime the focuser drawtube initially draws out (away from the telescope), focus compensation is invoked. Yes, it has to do with gravity. It’s possible that you didn’t set the backlash compensation value enough and it’s ok (and it’s better) to set it to large enough and overshoot. I set my compensation to same as auto focuser step size. It may be an overkill but but should work well.
My Auto Focus step size for 3.5" FeatherTouch focuser for TEC 140 APO is 100 or 125 steps so I set the backlash compensation to 100 or 125 steps and total of 9 steps. It’s better to overshoot backlash compensation to ensure that the compensation will work.
If you are not getting good “V” or “U” curves during AF run, try incrementally increase AF step size and don’t forget to increase backlash compensation to the same as AF step size. Or you may need to set backlash compensation to larger than AF step size.
Also, I like to set the step size and total number of steps so that the maximum HFR is about 2.5X greater than minimum HFR during AF run. Several others recommend at least 3X but I like less than 3X better.
What brand focuser are you using? Also what brand telescope are you using?
Peter
Hi Peter,
sorry for bothering you or the others, and thank you for your effort of explaining!!
My problem is as follows: I have NO (!) gravity backlash! It is a gear backlash! It happens any time you switch directions of movement. The value for both directions is the SAME!
The backlash comes not from the focuser, it comes from the stepper motor.
And I am not alone, there are plenty of guys having this problem, like the topic starter some years ago. And it seems like SGPro is not adressimg that problem.
APT has adressed that problem (!) but the autofocus routine is bad. So Id like to switch to SGPro but cant as there is no way to compensate gear backlash from the motor…
I have a USB-motor-foc V3…
Cheers
Chris
Interesting. Maybe contact USB-motor-foc V3 about your experience.
I am using Optec/Starlight Instruments Handy Stepper Motor with Focus Boss II focus hub and have not experienced backlash compensation issue with SGP.
It may not make much difference but can you tell us your brand scope, aperture size and focal ratio as well as brand of focuser?
Also, I am curious as why APT AF routine is bad? What does it do that is so bad and how can it be improved?
Peter
…first image of the stepper motor device, second of the APT gear backlash feature APT has…I wish SGPro had it :-((( So as you can see you can define both, IN and OUT, + you can define the gravity backlash in addition!
APT autofocus routine: uses only one star, SGPro uses many stars algorithm and shows the evaluation in a V curve. APT does only do a HFD on one star, much less convenient…
Thanks for the photos. You can try to post your message under Feature Request for backlash compensation for both directions.
Peter
Most ASCOM drivers for focusers have backlash compensation built in. You should not need to use the backlash compensation feature of SGP. When backlash compensation is used within the ASCOM driver, it will be implemented anytime the focuser changes direction. So, turn it on in ASCOM and off in SGP.
The same is true for temperature compensation – enable it in the ASCOM driver; not in SGP.
Charlie
1 Like
Hi Chris,
I’ve also seen what you describe where the reduction gears in the stepper motor mechanism cause this backlash. I have set the focuser up on a test bench with a dial gauge and measured the movement directly (ie not relying on star HFD measurements and V curve shapes) in and out and with direction reversals.
One option you might be able to look into is a different stepper motor controller that has backlash compensation in the driver. I have the Sharpsky stepper controller and I believe this has backlash compensation configurable in the driver - you could check with the developer Dave. That way you could disable the SGP compensation and configure it in the driver. I know it is not ideal to have to spend more on this and it would be better if SGP accommodated this form of backlash compensation but at least it is a potential path forward.
Peter
Optec’s FocusLynx focuser software has backlash compensation but it’s not clear whether it’s for both directions. Note “Reverse Directions” but it’s not clear if it’s for backlash compensation or simply for reverse spinning the motor. If you use Optec’s backlash compensation feature, then you must disable SGP’s backlash compensation feature. See image below:
Peter
OK. I just finished playing with Optec dedicated software and it’s not exactly great. With “Reverse Directions” disabled, the “OUT” direction moves the drawtube away from the scope. With “Reverse Directions” enabled, the “OUT” direction moves the drawtube toward the scope.
With backlash compensation enabled and “Reverse Directions” disabled, backlash compensation is invoked when the drawtube moves away from the scope. With backlash compensation enabled and “Reverse Directions” enabled, backlash compensation is invoked when the drawtube moves toward the scope. I do not see backlash compensation working in both directions at any time.
Peter
Hi Charlie,
in my case the ASCOM backlash compensation only allows one direction :-(.
Tried it aöready and ut really applies only to in or out, depending on what you set…
…thank you guys for checking!
Have great sunday and clear skies to you all!
Chris
You probably don’t understand how backlash in SGP works. It works by ensuring that the final focus movement is made from the same direction every time. If the move is in the backlash direction it is just made but if the move direction is in the opposite direction then two moves are made, first to a position past the final position, then a final move to the correct position. The amount of backlash made is not critical, all it needs to be is larger than the backlash in the system. In your example of 20 - 25 steps then something more than that will be needed, 50 steps for example.
The APT method is, I believe, not as good because it depends critically on the user setting the backlash corrections exactly correctly AND that the corrections are constant. As your post says your backlash isn’t, it varies between 20 and 25 steps. The corrections may need correcting and tuning as things change such as temperature.
The SGP method is much simpler, one number that is not critical, all it needs to be is larger than your mechanical backlash.
BTW, i wouldn’t be totally confident that your system is not affected by gravity. The camera and filter wheel hanging on the focuser could be enough to cause the camera to slip down, especially at high elevations.
…that is exactly the problem: I dont understand how SGPro does backlas compensation. It is not described for dummies like me, and still from your explanation I dont get it x-)
Do I understand you correctly: lets say the beginning movement is IN, we move 100steps. Then the direction changes to OUT 40steps. The gear backlash is 20steps, but SGPro does not compensate that when shifting to OUT, so although the motor makes 40 steps, only 20 are applied to the fucuser.
When I understand you correctly, SGPro then finishes the OUT movement by another IN Shift, applying IN backlash (of lets say 20, as set in the preferences). So the final outcome would still be: I wanted to move 40 out, only 20 out where applied to the focuser, and the backlash compensation for the next IN movement has been applied already.
…that still makes 20 missing…
On the gravity topic: sure you are right, that gravity makes an only IN backlash necessary, but that is right now the smallest of my problems (I guess it is in the one digit step range) as the gear backlash for every direction change is 20 steps. So that is a huge error compared to gravity backlash…
I agree, you don’t understand about backlash. I’ll make one more attempt.
Let’s imagine that you need to move to focus position 1000. The backlash is set to 100.
If you start at 500 then the move is done in two moves, first a move to 1000 + 100 = 1100. When that completes the final move to position 1000 is made.
If you start at 1500 then you also move to 1000 + 100 = 1100 followed by a final move to 1000.
In both cases the final move is in the same direction and by the same amount so as long as the mechanical backlash is the same and smaller than 100 the final position will be the same.
That’s all there is to it. The backlash correction is set to a value that is larger than any mechanical backlash and this means that the final move to the target position is always from the same direction and by the same amount so the mechanical position will be the same regardless of where the original position was.
…thank you Chris for your effort!! And I am sorry, i am too stupid, I still dont get it 
…I always thought I am not, but now there is proof! I have a PhD in chemistry, and now I dont know how I got that x-).
One last comment: well I tried it out in SGPro, I entered the backlash value and marked it as IN in the preferences tab. I shifted directions of movement in SGPro and the software did only compensate the error IN. So the next time movement to out I was lost. Thatis proof that it does not work…
So even if I am just too stupid to understand, SGPro fails…with my focuser. APT is doing not. Even if I enter larger backlas values, it gets messed up every time I see the ficuser moving/not moving/moving too short…
I guess we talk about different things: the process you describe compensates baclash coming frim the focuser. The backlash that bothers me is the backlas of the stepper motor alone…that is completely different…