AAF3 focuser stability?

[andrewlockwood]

Are there any other users out there with the new AAF3 focuser? I have one on my 16", and although I have no trouble connecting to it, I'd like to hear from others about how theirs behave in practice.

 

I have problems using it with my sequencing software (SGPro). When SGPro sends an ASCOM request to tell the focuser to move to a position, the AAF3 will move to a position within 1 or 2 microns of the requested position and stop. The focuser then reports it's slightly wrong position, and this figure is constantly flickering back and forth a couple of microns. Unfortunately, if this position isn't exactly what SGPro told it to go to, the software complains that the focuser isn't listening, and stops the sequence. This probably occurs 10% of the time, and always after I've left the software running on a sequence.

 

The authors of SGPro have refused my request to add in a focuser tolerance setting to their software, as they feel that a focuser should move exactly where it's told to, and then stay there, and if they tried to accommodate every hardware peculiarity they would not have a working product.

 

Although I don't want to make comparisons between the ASA Sequence software and SGPro, anybody who has used both will understand why I'd like to retain SGPro as my main observatory control software.

 

So can anyone else with an AAF3 tell me if this constant position change is normal, or should I expect greater stability from it? So far, repeated pleas to support have gone unanswered.

regards,

Andrew.

[LaCebra]

Hello Andrew,

I use an AAF3 since March, 2017. With ASA 12N.
Mine also oscillates between several values, of some microns, permanently.
I am still novice, thus I cannot know if it is a problem, or no …

 

On the other hand, when I make an absolute positioning: 6.620mm it works well (but with then an oscillation).
When I make a relative movement, with ACC : +/-20 microns for example, and which I begin again several times, the result moves gradually. For example: 6.639mm(instead of 6.640mm), 6.658mm(instead of 6.660mm),  6.676mm(6.680mm), 6.695mm(6.700mm), etc.

Then ,I have to make an absolute positioning in: 6.670mm!!!

 

I haven't still credibility to assert it more.  
La Cebra

[andrewlockwood]

Hello Andrew,

I use an AAF3 since March, 2017. With ASA 12N.

Mine also oscillates between several values, of some microns, permanently.

I am still novice, thus I cannot know if it is a problem, or no …

 

On the other hand, when I make an absolute positioning: 6.620mm it works well (but with then an oscillation).

When I make a relative movement, with ACC : +/-20 microns for example, and which I begin again several times, the result moves gradually. For example: 6.639mm(instead of 6.640mm), 6.658mm(instead of 6.660mm),  6.676mm(6.680mm), 6.695mm(6.700mm), etc.

Then ,I have to make an absolute positioning in: 6.670mm!!!

 

I haven't still credibility to assert it more.  

La Cebra

Thanks - I can understand why it might be hard to hold a Crayford style focuser within a micron, but this 'upgraded' focuser means my entire observatory is effectively useless for mosyt of the time as it cannot run unattended. I have yet to get sequence to take more than 4 images in a row before either it or Maxim crashes, so switching to another sequence program is not really n option.

My 25 year old JMI crayford and robofocus used to do exactly what it was told, so I have a choice between putting my old 10" newtonian back in the observatory, or buying another $3000 focuser from moonlight or feather touch so I can use my 16".

[MarkS]

Hi Andrew,

 

I use the AAF3/ASA10N with ASA Sequence. The reported position can vary by 1 - 3 micron during a run, but this is totally immaterial with either f/3.6 or f/6.8 configurations. I'm not sure what these small shifts mean, but the encoders are very precise, and I suppose they may be reporting very minor flexure - which is present in all systems.

 

Notwithstanding, the whole system is so stable that I have imaged for over 6 hours without using AF and not lost a single subframe. I could not do this with the OK3 although it did not report any shifts.

 

Mark

[andrewlockwood]

Hi Andrew,

 

I use the AAF3/ASA10N with ASA Sequence. The reported position can vary by 1 - 3 micron during a run, but this is totally immaterial with either f/3.6 or f/6.8 configurations. I'm not sure what these small shifts mean, but the encoders are very precise, and I suppose they may be reporting very minor flexure - which is present in all systems.

 

Notwithstanding, the whole system is so stable that I have imaged for over 6 hours without using AF and not lost a single subframe. I could not do this with the OK3 although it did not report any shifts.

 

Mark

Hi Mark, I only had one night to test the 16" a few months back to see how well it held focus, and it wasn't as stable as yours, requiring refocusing every hour due to both collimation shift and thermal changes. Now I've got the system together in one place again I'll try again as I seemed to have eliminated a fair bit of the collimation shift by tightening the spider a little.

I agree that 1-2 um does not constitute a significant change as the critical focus zone for this scope is about 30 um - but it's the impact on the sequence software that is doing my head in. Interesting to hear that you're getting by with sequence, so maybe it's worth chasing down whatever's causing the crashes.

[LaCebra]

Hi Andrew,

 

Your problem does not still exist for me.

 

I begin at present.

 

But I understand:

 

- Sometimes my camera refuses to connect to Maxim DL (I have to go out, switch off it, relight it);

 

- Sometimes Team Viewer changes the code of identification in laptop computer (I have to go out, read the new code, return to home);

 

I could not thus work with an observatory, but you, cannot go to sleep

I can !

 

Because when everything is settled, the images follow each other quite the night.

 

But the quality level for which I wait is much lesser than yours (at the moment).

 

 

A solution would be to insert a program inside the flow ASCOM (to read the flow - to modify it - to re-emit(re-utter) it modified).

 

I do not yet know how to intercept this flow.

 

If somebody know how ????

Then it would be possible to stabilize the value of the focuseur in a range of some microns.

If somebody knows how to intercept the flow ASCOM (with an example) I am inclined to think about it.

 

La Cebra

[RamaSpaceShip]

Hi All,

 

I don't know how to intercept the Ascom flow, but it is possible to implement a pseudo Ascom focuser that redirects the requests to the real one, gets its position, filters it and returns a stabilized position to the caller.

The application will connect to this pseudo focuser driver in place of the real one, and by that trick, will get fail-safe positions.

 

Hope this helps.

 

Bernard

[andrewlockwood]

Hi All,

 

I don't know how to intercept the Ascom flow, but it is possible to implement a pseudo Ascom focuser that redirects the requests to the real one, gets its position, filters it and returns a stabilized position to the caller.

The application will connect to this pseudo focuser driver in place of the real one, and by that trick, will get fail-safe positions.

 

Hope this helps.

 

Bernard

 

Thanks Bernard, a dummy focuser driver might be all that is needed. It would be simple enough, an integer division by 2 or 3 in each direction would probably do the trick. Unfortunately for me, the last time I was doing a lot of programming Fortran 95 was still pretty new and cool. I have absolutely no idea how to write an ascom driver, but I guess I will be learning.

Hi Andrew,

 

Your problem does not still exist for me.

 

I begin at present.

 

But I understand:

 

- Sometimes my camera refuses to connect to Maxim DL (I have to go out, switch off it, relight it);

 

- Sometimes Team Viewer changes the code of identification in laptop computer (I have to go out, read the new code, return to home);

 

I could not thus work with an observatory, but you, cannot go to sleep

I can !

 

Because when everything is settled, the images follow each other quite the night.

 

But the quality level for which I wait is much lesser than yours (at the moment).

 

 

A solution would be to insert a program inside the flow ASCOM (to read the flow - to modify it - to re-emit(re-utter) it modified).

 

I do not yet know how to intercept this flow.

 

If somebody know how ????

Then it would be possible to stabilize the value of the focuseur in a range of some microns.

If somebody knows how to intercept the flow ASCOM (with an example) I am inclined to think about it.

 

La Cebra

Thanks - I agree, it's all about being able to sleep. And I suspect you are being very modest about your own images, Do you have an astrobin account or similar?

[RamaSpaceShip]

Hi Andrew,

 

The code of an ASCOM focuser can be found here:

- https://sourceforge.net/projects/sglfocuser/files/Ascom%20Driver%20Software

- https://github.com/sirJolo/ascom-jolo-focuser/tree/master/VisualBasic/JoloFocuser

You can use them as an example.

 

Good luck.

 

Bernard

[andrewlockwood]

Hi Andrew,

 

The code of an ASCOM focuser can be found here:

- https://sourceforge.net/projects/sglfocuser/files/Ascom%20Driver%20Software

- https://github.com/sirJolo/ascom-jolo-focuser/tree/master/VisualBasic/JoloFocuser

You can use them as an example.

 

Good luck.

 

Bernard

Thanks Bernard - I've got the ASCOM developer's kit, the MS Visual studio installed, and am already working through the ASCOM example focus driver. Maybe in 10 years time I'll have figured out how to call an ascom driver from within another ascom driver.. Hopefully ASA will take pity on me before them and release a new driver for this focuser that solves all my problems. I'm also quite bemused by the wireless electronics in the focuser that don't do anything yet... I wonder when these features will be released to their owners?