AM5 Mount Best Practices

120 views
Skip to first unread message

Tim Stevenson

unread,
Dec 23, 2025, 1:27:03 PM12/23/25
to Open PHD Guiding
Attached is the log file from an evening operating my AM5 mount.  It includes a calibration and a Guiding Assistant.  The first section saw guiding of 0.96" Total RMS ... the second section saw that increase to 1.58" Total RMS.  I then ran Guiding Assistant and the guiding improved to 0.85" Total RMS.  Better but not where I think it could be.  I have been looking for recommendations for AM5 mounts and seem to have conflicting recommendations.  

A summary of the settings from Nov 22 PHD2 are:
Guide Exposure-     3s
Guide Rate- 0.5x
Dec/Ra Max Duration-  2500ms
Dec/Ra Aggressions- 100%/60%
MinMo- 0.18 / 0.12

On CloudyNights I found a posting from Kevin_A with the following recommendations:
Guide Exposure-     1s
Guide Rate- 0.5x
Dec/Ra Max Durations-    500ms
Dec/Ra Aggressions-   50%
MinMo- 0.1

Brian, several times in the forum I have seen you recommend very aggressive guiding for strain wave mounts which seems to conflict with Kevin's recommendations.

What are your recommendations for strain wave mounts and in particular the AM5?

Thank you
Tim Stevenson

Brian Valente

unread,
Dec 23, 2025, 1:48:52 PM12/23/25
to open-phd...@googlegroups.com
Tim

Brief comment:
>>>>Brian, several times in the forum I have seen you recommend very aggressive guiding for strain wave mounts which seems to conflict with Kevin's recommendations.

I seem to have created some confusion about this. By aggressive I primarily mean guiding cadence: short exposures times of 0.5-1 second. This is contrasted with a well behaved encoder mount that may need guiding cadence of more like once every 10 seconds.





--
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 visit https://groups.google.com/d/msgid/open-phd-guiding/662a966e-8a8a-4240-8059-582e6b2e3f4bn%40googlegroups.com.


--
Brian 



Brian Valente
Message has been deleted

Bruce Waddington

unread,
Dec 24, 2025, 11:55:02 AM12/24/25
to Open PHD Guiding
Hi Tim.  Here's what I suggest:
1. Update to the current version of PHD2 - the release you're using is nearly 5 years old
2.  Create a new profile using the new-profile-wizard.  Leave the guiding parameter settings alone
3.  Use the Calibration Assistant to get a fresh calibration in the preferred part of the sky.
4.  Set your guide exposure time to 1 second
5.  Staying in the same part of the sky, run the Guiding Assistant for about 15 minutes
6.  Stop the Guiding Assistant and let the system guide for another 15 minutes

Upload your logs and we'll take a look at how your system is performing.

Good luck,
Bruce

Tim Stevenson

unread,
Dec 25, 2025, 10:27:44 AM12/25/25
to open-phd...@googlegroups.com
Hi Bruce... I attempted to carry out your suggestions.  I updated PHD2 to the latest version and reloaded my equipment using the wizard.  I then ran calibration followed by ~15min of Guiding Assistant and ~15min of guiding.  I repeated calibration 3 times and here is what I did....

1) Initial setup of my mount including polar alignment.  Calibration using Calibration Assistant on the east side and target was on east side.  Ran Guiding Assistant for 1000sec.  Guiding was 0.87" for the 22m41s session with a polar alignment of 2.8' to 3.3' (Log sections 1-5).
2) Completed another polar alignment.  Repeated calibration using Calibration Assistant on the west side and target was on the east side very near meridian.  Ran Guiding Assistant for 909sec.  Guiding was 1.28" for the 27m59s session with a polar alignment of 0.2' to 1.8' (Log sections 7-12).
3) Slewed to a target on east side and removed my calibration, thus forcing a calibration while on target.  Ran Guiding Assistant for over 900sec.  Guiding was 1.12" for the 21m30s session with polar alignment of 0.5' (Log sections 14-17).


There is another difference between this date and the previous date... I updated the AM5 firmware and ASCOM drivers.  I'm making note of this because in some of the guiding sessions I see very large corrections but the guiding doesn't seem to be influenced by the spikes, in my opinion.

The first GA Result had me change Dec min-move to 0.18 and RA min-move to 0.11.  Backlash was 3245 and no recommendation was provided.  The second GA Result provided the same min-move recommendations but it did provide a backlash recommendation and I accepted it.  The third GA Result provided similar min-move recommendations and a different Dec backlash recommendation and it was accepted.

Tim

Brian Valente

unread,
Dec 25, 2025, 11:17:26 AM12/25/25
to open-phd...@googlegroups.com
Hi Tim

nice job on those logs and writeup. Your zip file did not include the debug log? 

Do you have a couple pictures of your setup from different angles? i'm interested in seeing how the guide scope is setup

Describe your overall system including imaging telescope. What is your image scale on your imaging scope? 

Some observations/suggestions:
- Your segment 5 results are in the ballpark of what i'd expect. Your min move translates to a little over 1" so guiding around 0.87" total RMS is expected. If you are aiming for lower you should try lowering your min moves. However your imagescale  of 6.45"/pix on your guiding setup is pretty coarse so you may be approaching the limits here.

- Enabling Dec Backlash degraded your results, so I would disable that.  

- in RA, you have a residual (guiding does not correct this) error at 62 seconds. This is likely something mechanical, perhaps a belt drive. It sounds like it's a known thing, so worth asking around in the zwo forums to see how people have dealt with it. 
image.png

Tim Stevenson

unread,
Dec 26, 2025, 2:01:50 AM12/26/25
to open-phd...@googlegroups.com
Thank you for your feedback Brian.  I was lucky and didn't tear my setup down.  I just picked it up and carried it to the garage (that's where I took a couple pictures, see attached).  That enabled me to just carry the rig out and place it in the same spot as the 24th.  So tonight I ran Calibration Assistant on the east side which is the same side as my target.  I also disabled DEC backlash.  I then ran GA for 1208sec (section 5) and the results were to decrease Min-Motion to DEC = 0.1 and RA=0.07.  Those values were lower than on the 24th and matches your recommendation.  Before the GA the rig was guiding at 0.9" (section 4) and after it increased to 0.99" (section 7).  However, I noticed Total RMS values of 0.5-0.6" while the rig was running.  Not sure what's going on there.  Logs for Nov, Dec 24th, and Dec 25th are attached.

I also checked the 62Hz content you mentioned and it's still there.  I will dig into the ZWO forums and try to find out what this is.

My rig ... AM5 mount sitting on the ZWO extension and a steel tripod.  This rig is using a William Optics Mini RedCat 51 as my main scope and a WO Uniguide scope (32mm / 120mm FL).  I'm using NINA and PHD2 of course.  I know the pictures are not the best and the cabling is also not ideal.  But since this was a temporary setup only to run the guiding test, I didn't spend a lot of time routing the cables.


Tim

Rig2.png
Rig1.png

Brian Valente

unread,
Dec 27, 2025, 9:37:30 AM12/27/25
to open-phd...@googlegroups.com
Hi Tim

your segment 7 run on 12/25 may offer some insight that your guiding issues seem to be transient

1. is your guidestar mass (snr shows something similar). It's a very steady signal
2. are points where your guiding is degraded, showing some oscillations particularly in Dec. total RMS is around 1.32"
3. When things are steady, your guiding is good - 0.62" total RMS and ra/dec having equal contributions

So I think your system is basically solid, you just have times when it experiences these oscillations. Your task is to hunt down the source of these. It could be wind, vibrations/environmental, axis balance, or even belt tension. 

Your Dec min move is still a bit higher and could probably be lowered to match RA. 

image.png

Reply all
Reply to author
Forward
0 new messages