Still problems with backlash comp with 2.3.13


#1

My Rigel Systems focuser is still having the same issue with backlash compensation in 2.3.13. I had two points where I had to manually nudge the focus to where SGP wanted it. The log is here:

https://dl.dropboxusercontent.com/u/58468743/sg_logfile_20140614212049.txt

I tried 2.3.5 the other night with no trouble, but 2.3.10 showed the problem. I’m running 2.3.7 right now and so far so good, but I’ve only done one autofocus run so far.

Kevin


2.3.14 and backlash compensation issue
#2

Well, the second AF run in 2.3.7 just failed the same way. Again, just nudging it manually got things back on track. I’m pretty sure 2.3.5 works, so I’ll go back to that for now.


#3

2.3.5 worked last night and tonight, handling AF runs with backlash comp with no trouble. So it looks like whatever went wrong happened in 2.3.6 (which I don’t have) or 2.3.7. Hope that helps.


#4

Ken or Jared,

Do you have a link to an archive for old SGP versions back to 2.3.5 or earlier?

Thanks,
Peter


#5

Well, I’m afraid I don’t have any interesting or great news here:

  1. We cannot reproduce this behavior in any test environment (so we are just guessing at this point)
  2. I pulled ALL of the auto focus code for 2.3.5 and 2.3.7 and there is not so much as a single byte of difference between them.

Sorry you are having difficulties, but I would not expect a fix for this anytime soon… Not sure where to look at this point.

Ken


#6

I should add a “thank you” for going through all the effort to pinpoint the issue here.


#7

Ken,

If you can provide a link to version 2.3.5, I would like to try it out myself to see if I am having similar issue.

Thanks,
Peter


#8

Kevin,

I think it would be better to ask Rigel to update their software to include backlash compensation in their ASCOM driver. Currently, their backlash compensation is tied to temperature compensation but not for a normal command to move or position the focuser commanded by SGP.

Peter


#9

Ken, the thanks go to you guys for trying to figure this out. Weird - I am having no problem with autofocus in 2.3.5. Peter, I asked the Rigel Sys folks for backlash comp in ASCOM, and they are working on it, but I think it might take a while. I think I"m going to have to stick with 2.3.5 for the near term.


#10

Peter,

We don’t keep old versions on hand. We would have to go through the code repositories and make a special build for you. Maybe Kevin would be kind enough to put his 2.3.5 installer on drop box for you.


#11

Rather I should say… we do keep old versions on hand, just not from that far back.


#12

I just downloaded and installed latest 2.3.13 and it fixed my issue. I am not sure why I didn’t try this version earlier. It looks like my issue was different than Kevin’s.

With focuser’s driver backlash compensation disabled and SGP’s backlash compensation enabled, I finally can send focuser’s new position to the focuser without SGP focuser’s locking up or stop communicating with the focuser. I can physically see the motor overshoot and then reverse direction to the proper focuser’s position. It never happened in previous two revisions.

Next test will be auto focus run. Hopefully the weather will clear up this weekend. If the auto focus run is unsuccessful, I will try SGP version 2.3.5. I still have focuser’s ASCOM driver backlash compensation that works well in case SGP 2.3.13 auto focus run with SGP’s backlash compensation enabled does not work.

Kevin, using SGP latest 2.3.13, have you tried a simple test like sending new focuser’s position and see if the motor overshoots and reverse direction to correct position?

Thanks,
Peter


#13

I tried 2.3.13, but get the same problem. The backlash compensation does run, so the motor backs up 500 steps and then reverses direction. About 2/3 of the time the focuser goes to the post-backlash comp position requested by SGP, and the AF continues as expected. But about 1/3 of the time the focuser stops anywhere from 1-10 steps short of the position requested by SGP, and then the sequence won’t continue.


#14

Hi Kevin,

Sounds like you might have to contact Rigel for a possible ASCOM focuser driver bug. It also sounds like a timing issue between ASCOM driver and SGP about deciding when the focuser motor should stop moving.

Like I said before, my next step is to do an Auto Focus run and see how consistent my ASCOM focuser’s driver is. I will make sure backlash compensation in ASCOM focuser driver is disabled while backlash compensation in SGP is enabled. Hopefully I will be able to test it this weekend.

Peter


#15

Kevin,

Is your backlash compensation really at 500 steps? What is your stepper motor setup resolution in microns per step? I am wondering if the timeout in either software is a little too short that’s causing to prematurely stopping the motor a little too early? Have you tried setting backlash compensation to less than 500 steps?

Rigel Sys web site says 0.1 degree per step and this translate to 50 degrees for 500 steps which seems quite large. What does 0.1 degree per step translate to in microns per step?

My motorized focuser is about 1.3 microns per step and I use 95 steps for backlash compensation.

Peter


www.mainsequencesoftware.com