RMS 4-6 !!! Is a Skywatcher SA GTI mount really this inaccurate?

[Walter ZAMBOTTI]

Equipment:

Mount = Skywatcher Star Adventure GTI. (with RA & DEC grub screws adjusted)

Camera = INDI Camera [QHY CCD QHY5III715C-d91] (with SVbony 190/50 guide scope)

Please tell me that it is something that I'm doing wrong!  After all I am new to PHD2 and guiding.

Log contains calibration and assistant information:

https://openphdguiding.org/logs/dl/PHD2_logs_s88b.zip

[Brian Valente]

HI Walter

Based on your longest run, the total RMS is less than that 

These are the main challenges I think you're facing:

First on the hardware side, the  overall periodic error is 15". PHD is able to guide it down to about 3" but i wouldn't expect a whole lot more from a PE that high. 

The second thing is you have some big instant jumps in your RA axis'  unguided results, sometimes 15" or larger. That is mechanical in nature,  you are going to need to look at the drive train:

On the operational side, you haven't really gotten a clean calibration, and you have probably gotten error messages to that effect if you haven't suppressed them. Make sur you use the calibration assistant, and follow its recommendations. The guiding assistant also pointed out your calibration needed to be redone, along with your guide camera needing better focus. 

The min moves you have are quite low for your setup, 0.5"? So you are hammering that mount with constant corrections. 

On the PHD side of things, for some reason you have dec auto backlash compensation disabled, so I would enable that for starters

You can try changing your RA algorithm to PPEC with a period of 453 seconds, disable auto adjust period. That should help improve things there.

Overall, you have some mechanical challenges and some cockpit errors, but I think you could improve things considerably by addressing the above.

What is your pixel scale of the imaging system?

Brian   

--
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/af18aa7b-78df-4901-8bda-a0c6fe72318an%40googlegroups.com.

--

Brian 

Brian Valente

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

[Walter ZAMBOTTI]

Brian. Thanks for looking at my issue.

The calibration assistant indicated I had orthogonal errors and suggested checking snagging cables unbalance weight setup and gear meshing (and polar alignment).  Cables and balance are good.  Polar alignment error was less than 1 arc-sec indicated by kstars.  The gears are beyond me.

The main scope image scale is 900/114 with APS-C sensor about 1.189 arcs/pix.

I have attached a screen shot of the calibration assistant:

I will try you recommendations as soon as possible.

I did tighten the mount gear grub screws to remove some very bad slack.  I trust/hope I didn't over tighten them.  I will have to revisit this

Regards

Walter Zambotti

[Walter ZAMBOTTI]

I have made the PPEC changes but cannot find where I can change the dec auto backlash compensation setting.  Is that in PHD2 or kstars or indi?

On Monday, January 27, 2025 at 10:32:10 AM UTC+8 bval...@gmail.com wrote:

[Brian Valente]

Your dec backlash is quite high and causing issues in your calibration 

On this graph, you can see the green (reversal) never actually reverses direction. That is likely the culprit and where I think you should spend some time adjusting your axis. You can adjust, test the calibration, adjust, and iterate.

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/e4161154-8897-4ce7-8223-ea168faca389n%40googlegroups.com.

[Brian Valente]

It's under the advanced settings (aka 'brain button') - here are some example settings, but as I mentioned before, you really need to get the backlash to respond. this can help but only so much

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/c2911a80-6d91-4d03-a1f1-99843c39a8b3n%40googlegroups.com.

[Walter ZAMBOTTI]

Not shown under my advanced settings!!!

[Walter ZAMBOTTI]

However I found this in the INDI driver for EQmod.

This shows the RA worm period as 474.  I don't know where the driver gets this value from.  Could the driver read it from the mount?  Should I use this value in the PHD2?

[Brian Valente]

Hi Walter

I'm not too familiar with the Indi implementation of PHD, but it may be dec backlash is not implemented there

Regarding EQMod, please avoid using any driver settings to improve backlash for now. 

My primary recommendation is getting your dec mechanical backlash improved. At this point it really doesn't matter what the software does or doesn't do.

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/198289a9-aec7-4985-a186-d6e1bac63186n%40googlegroups.com.

[Walter ZAMBOTTI]

I tried again (calibration and guiding)  last night with the RA algorithm set to PPEC with a period length of 453.   This did improve the results.  It drifted back and forth slowly from 3.2 to 1.5.  However the peaks were large.

I could see at times that the mount was trending east or west and PHD2 would give the mount commands to go the other way continually but the mount would take a long time to react.  This smells like slack.

As to the values in the INDI driver.   I didn't want to change the worm period values in the  driver (as they are read only anyway) but what I was curious about was using the 474 worm period value displayed in INDI to use as the PHD2 PPEC period length.

For now I have left it at 453 as you instructed, but just wanted to know your thoughts.

[mau...@gmail.com]

Hi Walter,

I struggled as well to stabilize the SA-GTI guiding performance in Windows 11 Ascom environment.

I share here my experience hoping it can be of some help in your case as well. 

The two decisive steps I went through in improving the performance were:

1. In EQMOD there are no SA-GTI presets available, in fact SA_GTI is NOT included in the list of supported mounts. Nevertheless EQMOD identified automatically a wrong default (maybe EQ5, I do not remember exactly), and consequently I went through a lot of troubles, for weeks. Long story short: I migrated from EQMOD to Green Swamp Software GSS, which supports SA_GTI, and this was the fix.

2. My mount came out of the factory with a huge backlash. Tightening the contrast screws is definitively simple, and I recommend you to do it. You can have a look to these two videos below. Please look at both of them before operating, as the second is amending a mistake included in the first one (too much disassembling), yet they together are a good tutorial.

https://www.youtube.com/watch?v=_nZr6yJWqOo

https://www.youtube.com/watch?v=4MdBxNvSDZ0

After this two fixes guiding became much more stable and reliable, with RMS in the range 1.0 to 1.7 arcsec, that is perfectly acceptable with my image scale of 5 arcsec. Ortogonality was always better than the one you posted above, yet it improved after the treatment.

Hope this may help.

Ciao,

Mau

[Walter ZAMBOTTI]

Thanks Mau

I have already tightened my grub screws for RA and DEC.  I am familiar with both those videos. Yes it made a big difference.  Hopefully I haven't over tightened them.  I would like to get my RMS down to 1.0 - 1.7!  I haven't lost hope

I have looked at GSS but it isn't supported on Linux, so isn't an option for me.

[Brian Valente]

Walter

I don't know if I would trust those numbers, i'm not sure where they come from via Indi. You might look up the spec from skywatcher on the worm period for the RA. Either way I'm not sure it will make a whole lot of difference. The number I gave you was the measured number from PHD.

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/7a34dcb3-4be9-42e3-8d01-e554c86da252n%40googlegroups.com.

[Bruce Waddington]

Hi Walter.  I don't think you can afford to ignore Mau's input, he may well be delivering inconvenient truth.  If the EQMOD driver doesn't natively support your mount, you may be stuck unless you can learn what settings to change.  I suggest going back to the absolute basics and using the 'Star Cross' test tool in PHD2.  It's described in the User Guide.  If you can't consistently get reasonable-looking cross patterns using this tool, there's no point in plowing ahead trying to guide.

Good luck,

Bruce

[Walter ZAMBOTTI]

Tonight should be clear.  I will perform the star cross test!!!

[mau...@gmail.com]

Ciao Walter, 

SA-GTI can definitively work with EQMOD, yet the correct mount parameters have to be set. I have been asking in the EQMOD community and you can find the conversation here. By the way, the linked chat includes the reference to the Chris Shillito tutorial indicating where to load the personalized mount parameters (for your comfort, see picture below).

I gave up because I could not find any complete answer or documentation about the parameters. I confess I have not tried to ask Skywatcher support, that sounds to me the most reasonable step forward. For me switching to GSS was a simpler path, unfortunately not viable for you in Linus environment.

Let me add one remark, please do not take me bad:  I would not work with periodic error correction before settling the correct base tracking parameters. To tell the truth, PHD2 manages so well the periodic error that I am not using PPEC with my SA-GTI, and on all my other mounts as well.

Hope this helps. As I said, I have been struggling a lot to have this mount work, and the reason is lack of documentation. I am surprised that the amateur community has neither found a solution nor discussing/documenting these issues with EQMOD (or maybe I missed something big?).

Good luck! It would be nice to hear about your progress.

Ciao!

Mau

[steve]

Hi
No need to struggle with INDI EQMod. The SA Gti has it's own INDI driver:

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/18354098-5056-495b-8b22-368c3d89f8ebn%40googlegroups.com.

[mau...@gmail.com]

I reckon this will be great news for Walter!

Steve, out of curiosity, is the INDI driver providing legible steps/full-sky-rotation and steps/worm-rotation? These values are not to be found anywhere! Thanks!

[steve]

We don't have a sa-gti ATM, but perhaps Walter could send you (a) screen-shot(s)? When connected, all the step/pointing/guide information is given in the INDI control panel.
Cheers and CS

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/7b3b7c67-50f9-4f79-969f-0affce2e77d5n%40googlegroups.com.

[Brian Valente]

nice!

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/92ed909e-66da-4801-b5d4-3c1012fd74af%40gmail.com.

[Walter ZAMBOTTI]

I tried changing the mount from EQmod to Star Adventure GTI (this is new).  But nothing works at all and PHD2 refuses to connect to that device.

[Walter ZAMBOTTI]

Here is the disappointing star cross 

Here is my star cross test when using EQmod.  The very short vertical line is the RA the long horizontal line is the DEC.  

At the moment I cant do the test after changing the mount from EQmod to STar Adventure GTI

[steve]

What do the indi logs say?

Do you have the LATEST (stable) kstars and indi? Method here:
https://www.indilib.org/download.html

Not sure which OS you're using but you'll also need the latest PHD2. For e.g. Ubuntu, Pitrick has made a repo here:
https://launchpad.net/~pch/+archive/ubuntu/phd2

Have you set the communication baud rate? It's 9600 using an eqdir cable, 115200 via usb B to A cable.
HTH

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/f112f1ad-ec4a-46fa-9019-8eac8211da73n%40googlegroups.com.

[Brian Valente]

Walter I think this is further confirmation your dec axis has excessive backlash and needs mechanical adjustment

Notwithstanding the good suggestions from community members here, I think this is the area you should focus on.

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/5c57e584-f907-4481-9f9f-a8fa4a65956cn%40googlegroups.com.

[Walter ZAMBOTTI]

Brian wrote:

Walter I think this is further confirmation your dec axis has excessive backlash and needs mechanical adjustment

Don't you mean the RA axis?  In my Cross Star Test the vertical axis of the landscape image is the East West RA and the horizontal is the North South DEC. My camera is rotated 90 degrees.

The RA axis barely moves, is this backlash???  The DEC axis has the larger movement but doesn't return to the center so clearly has some backlash.

What should I adjust or change to address this kind of issue?

Are there backlash values in the software I can adjust or do I need to fix the mount?

[Walter ZAMBOTTI]

Stev wrote :

Do you have the LATEST (stable) kstars and indi? Method here:
https://www.indilib.org/download.html

Not sure which OS you're using but you'll also need the latest PHD2. For e.g. Ubuntu, Pitrick has made a repo here:
https://launchpad.net/~pch/+archive/ubuntu/phd2

Have you set the communication baud rate? It's 9600 using an eqdir cable, 115200 via usb B to A cable.
HTH

I have the latest and kstars from the instruction on that site and I have the copy of PHD2 from that PPA.

The new SW SA GTI indi driver is connecting as I can manually slew the mount but that is about all I can do.

The problem seems to be the Park/Unpark and Track-On/Track-Off buttons in EKOS are greyed out.

I can park/unpark the mount via the indi panel but EKOS/kstars isn't aware of that

I will hop on the INDI forum and create an issue there. 

Thanks everyone for suggestions

[steve]

On 30/1/25 2:44, Walter ZAMBOTTI wrote:
> I will hop on the INDI forum and create an issue there.

Don't forget to include the INDI log along with the post.
Cheers

[Brian Valente]

I suspect it's dec, from your prior results i think you're reading the graph incorreetly

At this point it's mechanical. I previously sent you the backlash graph that shows your Dec not reversing at all despite many attempts by PHD to reverse direction

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/52b24b4c-c122-4018-a601-e20984195e37n%40googlegroups.com.

[Walter ZAMBOTTI]

I agree the graph indicates the DEC is a problem.

However while watching the looping of the guide scope during the star cross test I could read the step directions that were been taken by PHD2 and could see the movements in the loop images.

When moving west/east I could not discern any movement (in relation to the cross) but when moving north/south I could immediate see the movement.  The difference was chalk and cheese.

So assuming the graph is correct (which I do) then the mount having apparent problems moving west/east.  Then the issue is that PHD2 may not be taking into account that the DEC step resolution
is only half that of the RA step resolution on a SA GTI.

So I am confused but I want to try and understand what is going.

Much appreciate you for reminding me of the other (graph) results.

[Walter ZAMBOTTI]

The other data that the DEC problem seems inconsistent with is that after 40+ minutes of guiding the DEC RMS was 1.2 +/- 0.2 but the RA RMS was 2.6 +/- 0.2.

So I would be happy if the RA RMS was the same performance as the DEC.  It could have been random chance the DEC was that good.

I am setting up a Windows 11 PC (clean build) with Nina, AScom platfom, PHD2, etc and just won't to eliminate any software issues.  If I get similar results I can focus
on the mount.

But even if the mount has bad backlash isn't there a way to mitigate this via the software?

[Brian Valente]

Walter please upload the relevant logs, we can't really comment on anything without seeing them

>>>> But even if the mount has bad backlash isn't there a way to mitigate this via the software?

Yes, we discussed this previously a number of times. PHD2 has auto dec backlash compensation. It's unclear to me if the Indi version you are using has this feature, you should hop over to the Indi forum and ask there.

Brian 

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/9bf07828-666e-4285-ae99-b0dd53c792fan%40googlegroups.com.

[Walter ZAMBOTTI]

Some goods news.

I spent the last two days settings up a Windows/NINA/PHD2/GSS system and the results are much better.

Total RMS ranging from 0.9 to 1.7.   I sat with the scope for two hours on an almost windless night.  The RMS would spike on a gust of wind.  I should expect that.

I did have to fiddle with the aggressiveness settings to stop saw blading but it didn't take to long before I had the RMS settle down.

The Windows version of PHD2 has DEC backlash compensation, and the Linux version doesn't but I didn't use it during the calibration test.

I have screen shot the advanced settings and will try them on the Linux/kstars/phd2 system.

Strangely the Windows system also failed the cross test with the same pattern, but passed the calibration test with no warnings!!!

So I think the mount is OK (Relieved).  I need to work on the Linux setup some more.

The windows system also helped me discover a setup error I have with my Gemini Focuser on the Linux system but that is for another forum.

Thanks everyone for the help.  I will post back with further updates.

[Walter ZAMBOTTI]

So I have made a number of changes and retested under Linux.

Very disappointing. It still doesn't pass the calibration test.

Apparently this is the same mount as with the Windows test.  Hard to believe.

And a comparison with some guiding :

 

I don't even have to tell anyone which is which!!!

One physical difference is the Windows system was all software running local on one laptop with all devices USB connected.  The Ubuntu system is split client/server over Wifi.

I could try and provide a closer and dedicated hotspot and see if the network is causing any problem.

Any other suggestions?

[Walter ZAMBOTTI]

I spot another difference.

Why is the guide speed 90% in Windows and 50% in Linux?  And where can/do I adjust this?

[steve]

Is the server a RPI? Is the server running INDI and the client in a separate box running kstars?

If so, you're gonna need a cable connection. The latency of the WiFi is a killer. 

One solution is to run both INDI and kstars on the same box and simply vnc into that box. That way, you've solved the delays and you can concentrate on the imaging session.

HTH

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/89bb4a63-79af-429d-b8f1-6498362399e7n%40googlegroups.com.

[Michael Waring]

Hi Walter

How good is your Polar Alignment ?

There are instances in the first guide graph of RA and Dec simultaneously heading off.

Although that is also classic Othog Error behaviour.

Also I believe a higher Guide Rate and faster 1 second exposures would give PHD2 more omph, to make corrections harder and faster.

But let's see what the Experts recommend.

Michael

Wiltshire UK

[Brian Valente]

guidespeeds are set in the mount's driver

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/642e7089-e1f7-4979-bbd8-e37e78ca3a9bn%40googlegroups.com.

[Walter ZAMBOTTI]

to: stev...@gmail.com

Server is Odroid SBC, runningINDI.  Client is Ubuntu laptop.  I will try direct connect ethernet cable.

to:

treasurerb...@gmail.com

Polar alignment was excellent on both systems.  The first graph is of the Client/Server Linux setup. The second graph is the standalone Windows. I thought the borders should have given that away ;)

I am trying to figure out where I set a higher guide rate,  I don't see a setting in either kstars or PHD2.  I'm not sure how this got set to 90% on the WIndows system.

[Walter ZAMBOTTI]

To:

bval...@gmail.com

Thanks.  From memory I think the guide rate was set in GSS on the Windows system.

There are some guide rate settings in the INDI driver.  But there only seems to be 25%, 50% 75% & 100%.  I will ask on the INDO forum!

[Walter ZAMBOTTI]

I found the guide rate adjustment in the INDI driver.  It is a configurable percentage.  The other guide rate I saw previously was the ST4 guide rate?!?!?

I changed the rate to 90% on RA & DEC and re-ran the test but the calibration still fails and the guiding is not so good.

Comparing the Windows PHD2 Calibration test to the Ubuntu I wonder why there are so many different values:

In Windows the camera angle is -131.7 but in Ubuntu it is -141.3.  Camera angle relative to what?

Next test : I will now eliminate the Wifi and directly cable the client to the server.

[Walter ZAMBOTTI]

Latest INDI update has fixed the Star Adventure GTI driver issue I previously reported here.  So some more testing!!!

[Walter ZAMBOTTI]

So the updated Star Adventure GTI INDI driver didn't make any difference to the issue.

I went to directly ethernet cable the client and server together only to discover my laptop doesn't have an ethernet port!!!

So I bypassed the server all together and attached all the devices directly to the laptop.

First time I managed to get a successful calibration with Kstars/INDI/PHD2.  Guiding seemed more responsive/reactive but I had to deal with bad wind as well.

Is looking promising!!!

Which means if I want the client server setup I need to figure out what the issue is

[Walter ZAMBOTTI]

Some success!

Totally windless night.  Setup was kstars-client with remote indi-server.

I couldn't get the calibration to pass.  I noticed that the RA West/East calibration was looking very good but the DEC North/South calibration was causing some movement in the W/E directions.

I went into the indi driver and enabled R/A & DEC backlash at 10microsteps (minimum)  and then repeated the calibration and achieved this:

The rates are close to expected and shape isn't perfect but not horrible either.  Even the orthogonality has significantly reduced.

(Note: I don't think the R/A backlash setting effectively does anything as it only occurs with a change of direction in R/A, and since R/A is always progressing I never saw the driver issue backlash corrections on the R/A.  I did see the driver issue backlash corrections on the DEC)

However when guiding the automatic backlash setting in the INDI driver was causing saw toothing on the DEC axis. Probably because PHD2 was now over compensating.  So I disabled it and managed to get this on the guiding:

RMS varied from 1.2 to 1.8 hovering around 1.4 for most of the time.

So enable DEC backlash in the INDI driver (minimum micro-steps) on calibration and disable it when guiding.

It would be nice to see DEC backlash compensation on the Ubuntu version of PHD2.

[Walter ZAMBOTTI]

It turned out that the backlash compensation options as well as other options I didn't know exist were not being displayed because I was using an incompatible desktop theme in Ubuntu.

A quick change of theme and I can see/use all PHD2 controls.