Need a little help/advice with my AM5

[J Lewis]

Just got my AM5.  Backlash is severe. I have 2 questions:

1. Would upping the guide rate (currently 0.5 sidereal) help with the lash?

2. Does the PPEC algorithm essentially use hysteresis until it "learns" the PE of a mount?

See my guide log attached, along with a couple of screenshots.

Thanks!

[Brian Valente]

Hi J

>>>1. Would upping the guide rate (currently 0.5 sidereal) help with the lash?

A faster guiderate is always preferable, but I don't think it will do anything for your backlash. Looking at your guidelogs, it's hard to tell what the actual guiding will look like because on several runs guiding is uni-directional (indicating polar misalignment). Were you adjusting your polar alignment in this run? In places where it does reverse, it doesn't look too severe. You might try enabling auto backlash compensation and see if it cleans it up. the dec and ra total RMS aren't that far apart.

The bigger issue I see is these random large spikes in Dec, 

Not sure where that comes from , but it's something to hunt down

>>>2. Does the PPEC algorithm essentially use hysteresis until it "learns" the PE of a mount?

Yes that's correct, although it starts with hysteresis and blends in the predictive part over the course of several periods

--
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/131d4733-9947-4e3f-98d4-400155ad15bfn%40googlegroups.com.

--

Brian 

Brian Valente

astro portfolio https://www.brianvalentephotography.com/astrophotography/

portfolio brianvalentephotography.com 

astrobin https://www.astrobin.com/users/bvalente/

[J Lewis]

Hi Brian,

No, I wasn't even outside with the scope.  I did my polar alignment before hand, came inside, got on my PC and started her up.  According to PHD2, I was at 1 arc minute of PA error, I'd assume that's decent.  Those jumps happen OFTEN.  Others with the mount I've spoken to on CN have stated they get the same thing, just not the terrible guiding that I get, lol.

[J Lewis]

Just to add, I ran thru 2 iterations of polar alignment using Sharpcap.  According to Sharpcap, I wasn only off by 7".  But I realize different softwares see things differently.

[Brian Valente]

From the data shown, it's clear it's guiding in only one direction, so there are three posssibilities

- polar misalignment (the most common)

- uni directional guiding is chosen as the setting 

- flexure, loose equipment, snags, or other gear shifting that causes guiding in one direction

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/af431b22-07f5-4634-9c7b-c2a28e202849n%40googlegroups.com.

[J Lewis]

Thanks for your help Brian.  There is another possibility.  A lemon mount, but I'm hoping not, lol.

Is this the setting you suggested?  It's already checked.

[Brian Valente]

No, that's different

Auto backlash compensation is in the algorithms tab (and also described in the documentation)

here it is along with some suggested default settings for you:

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/04bff747-dda2-4048-945c-10a30adf2efcn%40googlegroups.com.

[J Lewis]

Just one last question, where would I find the setting for Unidirectional guiding?

[J Lewis]

Thanks for that.  I've checked the box and applied those numbers.  All I need is a cloudless night...  I'll up the guide rate to .9 and see how things go.

This thing is very light, but my EQ6 is a MUCH better performer.  I'm hoping I can get it to work to save my back.

Your help is greatly appreciated, many thanks.

Jeff

[Brian Valente]

see where it says ded guide mode - auto is bi-directional

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/a09caa51-85df-434b-85fa-2b303dd719fbn%40googlegroups.com.

[J Lewis]

Gotcha.  Mine is set on Auto, so that's one possibility ruled out.

Thanks again!

[Mike Jerry]

I am watching this thread with interest as I am in the same boat as you with my AM5. I have contacted Agena about this who has sent a report to ZWO to see how they will react. Maybe a loose Dec axis belt?

I will have to say, that even with this massive lash, my guiding is actually pretty good. Last night I was between 0.35 and 0.7 all night.

I actually came here to see if anyone has figured out any PHD magic settings with these new mounts.

[Brian Valente]

Hi Mike

>>>I will have to say, that even with this massive lash, my guiding is actually pretty good. Last night I was between 0.35 and 0.7 all night.

Just curious, what expectations do you have for 'magical settings'? 

Assuming you mean RMS, those guiding stats should produce tight round stars. 

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/b7ed9f7f-d308-444c-9290-55790244b508n%40googlegroups.com.

[J Lewis]

Hi Mike, yeah, I'm on the same thread with you on CN.  I wish I could get your guiding!!  Those are great numbers.

[J Lewis]

Deleted my last comment.  I understand some things can be misread.

[Brian Valente]

I'm sorry if it came across that way - i certainly did not intend that

Usually when people ask for suggestions, it's because guiding is less than ideal

0.3 - 0.7 RMS would be considered good-to-excellent guiding by many, myself included

If dec backlash was a significant issue, I would expect to see larger numbers, but maybe I misunderstood.

That's why I asked what is the concern?

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/936de2a2-edce-4c3e-b0b6-9f587bf6b173n%40googlegroups.com.

[Mike Jerry]

Yes, I do get nice round stars. No complaints there. I guess I meant to say that I was wondering if any best practices had been established for strain wave mounts with PHD yet.

[Mike Jerry]

While I do get good to great guiding, there is one thing that continues to puzzle me. From one GA to the next, even on the same night, I can get WILDLY different Dec backlash measurements from PHD. To date, I have seen everything from 3000-6000ms! And I'm not alone. Others on the monster AM5 Cloudy Nights thread have also reported large swings in the Dec backlash being reported by PHD.

I don't believe I have seen such behavior from other 'traditional' mounts, but maybe I need a reality check. Is this common?

[Bruce Waddington]

The Guiding Assistant is actually measuring reversal-delay time which is not necessarily the same as mechanical backlash.  It’s usually affected by pointing position and sometimes by scope imbalance.  You can read more about it here:  https://openphdguiding.org/man-dev/Supplemental_Info.htm#Common_mount_problems

 

From what I’ve seen, many of these strain-wave drive systems can be pretty inaccurate, I think some of the manufacturers are happy with RA tracking error of 20 arc-sec.  And worse still, the tracking errors aren’t periodic or predictable so things like periodic error correction aren’t very effective.  I think this is because the changing points of engagement between the drive elements have their own, unique manufacturing errors.  I suspect this is why some of the newer products now include high(er) precision encoders in RA.  I know you’re asking about Dec but I would expect the inaccuracies would apply there as well.  In any case, I think the mounts may be less amenable to high-precision auto-guiding because of their lower repeatability and predictability.  I think one common practice is to use shorter guide camera exposures while relying on multi-star guiding to help avoid chasing the seeing.

 

Good luck,

Bruce

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/3bafd611-1adb-4ad8-a2bc-1868d142f03en%40googlegroups.com.

[J Lewis]

Thanks Bruce.  I think with the newness of the strain wave mounts (at least in the general population now), we're seeing things we're just not used to, because the machines themselves are very different.  When the mounts with an assumed backlash of 6000ms are getting .3" rms guiding, it's bound to draw questions.

In my case, I'm not getting that kind of guiding, but I've only had the mount out twice.  I'll do more testing, and if I can achieve the same results (anywhere below say .85" rms), then I'll be pretty happy.  Folks are generally happy with the mounts, and as a rule, they are guiding at the purported 0.5" to 0.8" rms without issue, with many performing even better than that.

Thanks to you and Brian for your input here!

Best regards,

Jeff

[Mike Jerry]

Thanks for this Bruce.

In terms of position and imbalance, I always have the scope in the same position on the mount (I have marks on the dovetail bar to ensure that) and I am also less than a degree off zero declination and within 2° or 3° of the Meridian when I calibrate and run guiding assistants. In sum, there is little to no change in either of these things and yet I am seeing radical differences in the reported Dec backlash.

As I said, I am getting very good guiding from it with PhD, so I don't think there is an actual problem with the mount itself. I think we are beginning to form a hypothesis that the "reversal-delay" method (though I don't technically understand what the difference is between that and actual backlash measurement) is not necessarily an accurate methodology for measuring backlash with strain wave mounts.

And, I am definitely using shorter guide pulses than I did with my AVX. One second seems to be about the sweet spot.

Mike

[Brian Valente]

Hi Mike

>>>(though I don't technically understand what the difference is between that and actual backlash measurement)

I think of it this way:

backlash is mechanical: The angle that the output shaft of a gearhead can rotate without the input shaft moving. It can arise due to tolerance in manufacturing and/or the gear teeth need some play to avoid jamming when they mesh

reversal delay includes backlash, but may also include other factors as well: software-based tuning (such as a software-based backlash compensation, periodic error correction, etc.), encoders, spring-loaded mechanisms, etc. 

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/8e3838e3-3e89-4276-97e3-0d713925ddd2n%40googlegroups.com.

[Mike Jerry]

Good points. Thanks Brian.

Mike