Backlash compensation sometimes fails to complete

Sequence Generator
1

I noticed last night that some frames were out of focus, apparently because backlash compensation didn’t run as expected. The logfile is here:

The log shows that backlash comp ran fine most of the time, but it failed to run correctly on some frames, seemingly at random. Green_frame67 was one of the out of focus images. The log shows that backlash comp ran 500 steps OUT as expected, but failed to move IN again as it should have, so the focus was off by 500 steps. Yet the next frame (Blue_frame65) was in focus, despite starting 500 steps out of focus. The log for that frame seems to show that backlash OUT ran as expected, as did backlash IN. That would have left the focuser 500 steps too far out, just like the previous frame, except that backlash IN then ran again, putting the focuser in the correct position. This issue cropped up seemingly at random during the session. Any idea what’s happening, and how to correct it? Focuser is a Rigel nStep.

Kevin

2

@astrovienna I took a look, but I didn’t see the behavior you describe. I did not look deeply so I apologize if I just missed something. For the out of focus frame (green_frame67), I extracted this from the logs:

[06/24/19 01:11:25.319][DEBUG] [Focuser Backlash Thread] Focuser backlash thread has started...
[06/24/19 01:11:30.333][DEBUG] [Focuser Backlash Thread] Backlash thread has detected that movement to modified position (36356) is complete...
[06/24/19 01:11:30.333][DEBUG] [Focuser Backlash Thread] Moving focuser to original requested position (35856)...
[06/24/19 01:11:30.333][DEBUG] [Focuser Backlash Thread] Focuser backlash compensation needed. Steps: 500 Direction:IN
[06/24/19 01:11:30.333][DEBUG] [Focuser Backlash Thread] Focuser moving to 35856
[06/24/19 01:11:30.334][DEBUG] [Focuser Backlash Thread] Focuser move call complete
[06/24/19 01:11:33.917][DEBUG] [Sequence Thread] Focuser position matches requested position (35856), continuing...

This looks pretty normal to me. Where are you seeing a failure to correct the backlash movement?

3

Thanks for looking at this, Ken. I may be misreading the log - everything has been going well for a long time, so I’m pretty rusty at troubleshooting. Alas the one thing I’m sure of is that frame 67 was one of the out of focus ones: SCT donuts. I assume the process for that frame begins with this line:

[06/24/19 01:08:43.122][DEBUG] [Sequence Thread] ------------- Starting capture frame for event[1] -------------

Then SGP rotates the filter wheel to put green in place and recognizes it needs a 500 step OUT move for backlash comp:

[06/24/19 01:08:43.384][DEBUG] [Sequence Thread] Old filter: Red; Focus pos: 35037
[06/24/19 01:08:43.384][DEBUG] [Sequence Thread] New filter: Green; Focus pos: 35067
[06/24/19 01:08:43.384][DEBUG] [Sequence Thread] Pos delta: 30 steps
[06/24/19 01:08:43.384][DEBUG] [Sequence Thread] New position-> (current: 35229) + (delta: 30) = 35259
[06/24/19 01:08:43.384][DEBUG] [Sequence Thread] Focuser moving to 35259
[06/24/19 01:08:43.388][DEBUG] [Sequence Thread] Focuser backlash active, modified move to 35759

I think the OUT move is completed a few lines later:

[06/24/19 01:09:44.095][DEBUG] [Sequence Thread] focuser move is complete…

The frame capture begins shortly after this, but there’s no corresponding 500 step IN move to get the focuser to 35259, where it should be for that filter. This seems to be confirmed because the next frame (Blue_frame65), which starts here:

[06/24/19 01:11:23.755][DEBUG] [Sequence Thread] ------------- Starting capture frame for event[2] -------------

reports that the focuser is at 35759, not 35259:

[06/24/19 01:11:24.014][DEBUG] [Sequence Thread] Old filter: Green; Focus pos: 35067
[06/24/19 01:11:24.014][DEBUG] [Sequence Thread] New filter: Blue; Focus pos: 35164
[06/24/19 01:11:24.014][DEBUG] [Sequence Thread] Pos delta: 97 steps
[06/24/19 01:11:24.014][DEBUG] [Sequence Thread] New position-> (current: 35759) + (delta: 97) = 35856

So it looks like for Green_frame67 backlash comp moved out, but it didn’t move back in. The really odd thing is that the next frame, Blue_frame65, is in focus despite starting 500 steps out of focus. The log seems to show that backlash comp ran 500 steps OUT and then 500 steps IN, as expected. But then it ran another 500 steps IN, which put the focuser into the proper position.

Reading logs is mostly way beyond me, so apologies if I’m misinterpreting this. Thanks again for your help.

Kevin

4

Ken, I think the section of log you put in your post was for Blue_frame65, which was actually in focus. It was the previous frame, Green_frame67, which was out of focus.The log sections I added above are for that frame.

Kevin

5

I have encountered probably the same problem. It appears ones so many exposures, and I have had it already in two different photo sessions where this happened. For me the focuser backlash compensation is set to 400 units, as this offset can be recognised in the logs. I have made a screen shot of the photo files with the focuser setting at the moment of exposure and highlighted where the error occurred. I have also added the SGP log of one of the sessions, but I could not spot the error there yet.
Attached the screen shot and log file: Link to 2 screen shots and 1 log file.

Best Regards, Mario

6

Hi Ken and Jared (and welcome to Mario!),

This continues to happen. Take a look at Blue_frame94 and Blue_frame95 in this log:

Those frames are defocused, apparently because the backlash IN move didn’t happen, but the red and green frames around it are in focus. And the blue frames before and after these two are also in focus.
Tonight I actually caught this happening live. The filter wheel dialog box says “changing” for a long time, and the focuser has already done the OUT move for backlash comp, but it stays at that position and doesn’t do the IN move. Finally SGP seems to give up waiting for the focuser to move, and proceeds with the exposure. The filter wheel dialog box says “moving,” and the filter is set, and the exposure begins.

So as not to be a total downer, here’s a pretty picture, all taken with SGP. Sh2-235 in Auriga:

Sh2-235 by The Hole in the Trees Skybox

Thanks for looking at this.

Kevin

7

Yesterday, since a long time, the weather was fantastic and I made about 60-70 subframes of the Iris Nebula and Stephan’s Quintet. As a containment I have added breaks between exposures (Delay First and Delay Between in the Sequencer window) of both 3 s. I rotate the filters through the exposure events. This can also be entered as a default under Tools/Options. Now, I did not find any backlash error, and all frames were in focus. I am using SGP 3.0.3.169, and it looks like this containment works. Hope this works for you too, and hopefully the root cause of the problem can be eliminated as well later in SGP (or Sesto Senso, who knows?).