Problems guiding with Dec jumps/backlash

337 views
Skip to first unread message

David Rawlins

unread,
Aug 17, 2023, 10:41:14 AM8/17/23
to Open PHD Guiding
I'd be very grateful for help diagnosing a problem with my tracking/guiding which is preventing me from imaging.

I have previously had successful guiding using my HEQ5-R mount and SW150PDS but last night trying to image M27 at about 60 degrees altitude, the guiding assistant was saying I had very significant Dec backlash of between 5 and 9 secs. What I was seeing while guiding was large jumps in Dec which PHD2 was then trying to correct. It did an ok job after I applied the assistant's settings but total error was way higher than I'd expect and not good enough for imaging.

The first thing I suspected was cable snagging or balance problems but I haven't been able to find any issues. I have adjusted the Dec worm gear on the mount several times and am pretty sure there isn't significant backlash there so I suspect there's something else going on.

The calibrations I did didn't throw up any errors but I've not yet tried a star test or nudging the mount North before guiding which I know might be useful. Can you suggest anything from the log file that would help work out what is going on?

Thanks very much in anticipation
David
PHD2_GuideLog_2023-08-16_211503.txt

Bruce Waddington

unread,
Aug 18, 2023, 4:31:26 PM8/18/23
to Open PHD Guiding
Hi Dave.  There are two significant problems with the Dec performance on the mount.  The first is that you *do* have a lot of Dec backlash or perhaps you've over-tightened the gear mesh already and have created binding on the axis.  It's also possible that something like a set screw has come loose since you made adjustments. But you can easily see how long it takes for the mount to respond to Dec guide commands when the direction of guiding switches (Dec is in green):

Dec_Backlash.jpg

Second, these huge guide excursions are not coming from the mount drive system or PHD2 - they happen spontaneously.  The arrow below points to the time when the excursion began and we know the Dec motor wasn't running and PHD2 hadn't issued a guide command to move north:

Dec_Excursion.jpg

This is coming from something that's moving around on your guiding assembly because of a loose fitting or cable pulls.  I think you probably need to internalize how small these movements really are.  The largest Dec excursion in the top graph is equivalent to an unwanted movement of the guide camera sensor by 36 microns, or about 50% the thickness of a human hair. 

These are some of the easiest problems to identify in the guide logs.  If you use the PHDLogViewer tool and review the log analysis tutorial, you can easily see what's going on.
Hope you can get things sorted out quickly,
Bruce

David Rawlins

unread,
Aug 19, 2023, 5:10:40 AM8/19/23
to Open PHD Guiding
Thanks very much, Bruce, you've given me some good avenues to investigate. 
I "hypertuned" my mount in June and while it runs very freely now, it seems harder to adjust the worm gear optimally and as you say, I may have some binding going on. 
My guidescope is on a simple finder bracket at the moment - not at all ideal and that may well be the cause of the movement you've detected. I won't bore this thread with the details but more to investigate.
Thanks also for the links to the viewer and tutorial
Best wishes
David

Bruce Waddington

unread,
Aug 19, 2023, 3:31:50 PM8/19/23
to Open PHD Guiding
Wheezy, "stalk-mounted" finder scopes are real bad ideas, almost a guarantee of flexure and unwanted movement.  Just keep thinking about the fact that your guide camera can "see" movements of 3-4 microns.

Regards,
Bruce

David Rawlins

unread,
Aug 23, 2023, 3:58:48 AM8/23/23
to Open PHD Guiding
Thanks again, Bruce. 

I mounted my finder on a dovetail and moved everything else around to achieve balance and I tightened up the worm drive gear and the drive belt which was a bit slack. I thought I'd cracked it, with your help, as the guiding assistant reported less backlash (but there's still some but it appears manageable) and the guiding itself seemed to be working well. But soon after, I started getting big fluctuations in Dec again. This time, they seem to be caused by large Dec inputs from the guiding.

See for example session 3 in the attached. It was initially very good, then I must have nudged something which caused a big jump but after that, the periodic large Dec inputs seem to be reacting to small Dec movements. Subsequent sessions were worse. I accepted the guiding assistant recommendations on min mov and repeated it a couple of times but it didn't seem to help.

So it looks like I've improved the flexure situation and the backlash (but more to do there) but can you advise how I can fix these large Dec inputs?

Best wishes
David
PHD2_GuideLog_2023-08-22_210750.txt

Brian Valente

unread,
Aug 23, 2023, 9:49:07 AM8/23/23
to open-phd...@googlegroups.com
Hi David

It looks to me like your backlash compensation amount is set too high, try starting around 800ms and it will auto adjust itself to the right amount. I know the guiding assistant said about 2350ms but based on your results that looks too high



--
You received this message because you are subscribed to the Google Groups "Open PHD Guiding" group.
To unsubscribe from this group and stop receiving emails from it, send an email to open-phd-guidi...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/7fefa1ef-ae97-4424-841f-a3112f07d7c1n%40googlegroups.com.


--

djra...@gmail.com

unread,
Aug 23, 2023, 11:28:31 AM8/23/23
to open-phd...@googlegroups.com

Thanks, Brian, will do

David

 

Second, these huge guide excursions are not coming from the mount drive system or PHD2 - they happen spontaneously.  The arrow below points to the time when the excursion began and we know the Dec motor wasn't running and PHD2 hadn't issued a guide command to move north:

 

 

This is coming from something that's moving around on your guiding assembly because of a loose fitting or cable pulls.  I think you probably need to internalize how small these movements really are.  The largest Dec excursion in the top graph is equivalent to an unwanted movement of the guide camera sensor by 36 microns, or about 50% the thickness of a human hair. 

 

These are some of the easiest problems to identify in the guide logs.  If you use the PHDLogViewer tool and review the log analysis tutorial, you can easily see what's going on.

Hope you can get things sorted out quickly,

Bruce

 

On Thursday, August 17, 2023 at 7:41:14 AM UTC-7 djra...@gmail.com wrote:

I'd be very grateful for help diagnosing a problem with my tracking/guiding which is preventing me from imaging.

 

I have previously had successful guiding using my HEQ5-R mount and SW150PDS but last night trying to image M27 at about 60 degrees altitude, the guiding assistant was saying I had very significant Dec backlash of between 5 and 9 secs. What I was seeing while guiding was large jumps in Dec which PHD2 was then trying to correct. It did an ok job after I applied the assistant's settings but total error was way higher than I'd expect and not good enough for imaging.

 

The first thing I suspected was cable snagging or balance problems but I haven't been able to find any issues. I have adjusted the Dec worm gear on the mount several times and am pretty sure there isn't significant backlash there so I suspect there's something else going on.

 

The calibrations I did didn't throw up any errors but I've not yet tried a star test or nudging the mount North before guiding which I know might be useful. Can you suggest anything from the log file that would help work out what is going on?

 

Thanks very much in anticipation

David

--
You received this message because you are subscribed to the Google Groups "Open PHD Guiding" group.
To unsubscribe from this group and stop receiving emails from it, send an email to open-phd-guidi...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/7fefa1ef-ae97-4424-841f-a3112f07d7c1n%40googlegroups.com.


 

--

Brian 

 

 

 

Brian Valente

--
You received this message because you are subscribed to a topic in the Google Groups "Open PHD Guiding" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/open-phd-guiding/7BVyr8kFMDo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to open-phd-guidi...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/CAJa45i55fqh5jb931Y92gkh%2B8DgfnU226ecbCpN_mQ_zVV_EDA%40mail.gmail.com.

~WRD0000.jpg

Bruce Waddington

unread,
Aug 23, 2023, 12:27:29 PM8/23/23
to Open PHD Guiding
Could you please upload both the guide and debug logs for this session using the built-in Uploader feature:

I suspect you've over-tightened the gear mesh and created stiction in the drive system.

Bruce

David Rawlins

unread,
Aug 24, 2023, 3:38:38 AM8/24/23
to Open PHD Guiding
Thanks again, Bruce, the files are here:


I really appreciate the time you are spending helping me with this. And I've learned a new word - stiction - and found out what it means!
Best wishes
David

Bruce Waddington

unread,
Aug 24, 2023, 12:50:29 PM8/24/23
to Open PHD Guiding
It looks to me like the Dec drive system is probably too tightly meshed and it's causing the backlash test to be overly pessimistic.  Here's what the test showed:

2023-08-24 09_40_55-Backlash Test Analyzer.jpg

The red dots are north guide pulses, the green ones are south.  You can see that the first two south pulses continued to show north movement in the mount - we can actually see the same thing in the calibration.  If we look at the history of how the backlash compensation was adjusted during your sessions, we can see it was consistently reduced down to about 1 sec:

2023-08-24 09_40_00-Backlash Comp Analyzer.jpg

You made life harder for yourself by running the backlash test in the GA in the middle of your imaging session.  There's no point in doing that unless you've made changes to the gear mesh.  When you did that and reset the backlash comp value, you started all over with this adjustment process.  Try setting the value at about 1 second and then leave it alone - don't reset it unless you have a reason to think something has changed.  In the Guiding Assistant, just un-check the box for measuring backlash.  I'm also making the assumption that you don't have any backlash compensation enabled in the EQMOD driver or in the mount firmware.

Regards,
Bruce

djra...@gmail.com

unread,
Aug 24, 2023, 3:55:27 PM8/24/23
to open-phd...@googlegroups.com

Thanks again, Bruce, that makes complete sense and I don’t have any backlash compensation in EQMOD or the mount. I’ll try backing off the tension on the worm drive a tiny bit and use your approach to setting the compensation if there’s still an issue.

 

I’ve seen some people use an ammeter to measure how hard the drive motor is working and using that to fine tune the tension. Hopefully not necessary in my case.

 

Best wishes

David

 

PS For any other relative newbies reading this, I thoroughly recommend looking at Bruce’s tutorial that he linked to in this thread as it gives real insights into what is going on during a guiding session and in interpreting the logs.

 

 

 

From: open-phd...@googlegroups.com <open-phd...@googlegroups.com> On Behalf Of Bruce Waddington
Sent: 24 August 2023 17:50
To: Open PHD Guiding <open-phd...@googlegroups.com>
Subject: Re: [open-phd-guiding] Re: Problems guiding with Dec jumps/backlash

 

It looks to me like the Dec drive system is probably too tightly meshed and it's causing the backlash test to be overly pessimistic.  Here's what the test showed:

 

 

The red dots are north guide pulses, the green ones are south.  You can see that the first two south pulses continued to show north movement in the mount - we can actually see the same thing in the calibration.  If we look at the history of how the backlash compensation was adjusted during your sessions, we can see it was consistently reduced down to about 1 sec:

 

Image removed by sender.

 

Second, these huge guide excursions are not coming from the mount drive system or PHD2 - they happen spontaneously.  The arrow below points to the time when the excursion began and we know the Dec motor wasn't running and PHD2 hadn't issued a guide command to move north:

 

Image removed by sender.

~WRD0001.jpg
image001.jpg
image002.jpg
Reply all
Reply to author
Forward
0 new messages