Unusual PHD2 Performance After OAG Configuration

[Juno16]

Hi PHD2 Forum,

I recently installed an off axis guider to replace my guide scope/camera.

I focused the main imaging camera and then the guide camera (ASI220mm mini).

I calibrated PHD2 and ran the GA.

I did some test guiding and the results were remarkable, too remarkable.

My guiding was 0.2-0.3 total rms error (arcsec) which is great, but it seems way too great.

When I was calibrating, I stepped back from my pc (indoors by remote desktop) and I saw what looked like four star image displays. Upon closer inspection, I can see that the star patterns in each of the four star image displays looks the same. 

Not sure if this is related, but most of the time PHD2 auto-selected a guide star that was saturated (flat top on the star profile) and the snr was high at 150-200.

Unfortunately, the only screen snip that I captured where the four star images can be seen (barely) was during calibration. Below is that snip. It is kind of dark, but to me it looks like four identical images (or panes). The star patterns look similar in each.

I also included a snip of PHD2 (unfortunately after I turned down the gamma slider) when guiding and a snip of the GA results.

This is a link to my uploaded log file. 

https://openphdguiding.org/logs/dl/PHD2_logs_vKEc.zip

Any thoughts what might be causing this? If it were simply duplicated images that wasn't affecting the guiding performance, I would be thrilled. 

Are the multiple star image panes limiting the ability of PHD2 to guide correctly and resulting in unusually "great" guiding?

Equipment: Skywatcher HEQ5/Explore Scientific ED102 (reduced to 571mm)/ASI533MCP camera

Thank you so much for the assistance!

Jim

[Juno16]

I forgot to include in my previous message that I did try replacing the usb cable to the guide camera.

[Brian Valente]

Hi Jim

I don't see any obvious errors in your guidelogs, although there are somewhat suspicious things. How do the stars in your subs look?

I don't see anything that would suggest you are guiding on a hot pixel, although the guidestar SNR is fluctuating quite a bit. Guiding on a saturated star should obviously be avoided and may give you false sense things are going better than expected, so just make sure your sat by max adu is set appropriately (256 for 8 bit or 65535 for 16 bit) or use saturation by star profile

First, just confirming you used a new profile via the profile wizard?

Your settings look okay, but the # of steps during calibration is abnormally large. I would expect about 12 steps, and instead you have 2x the number of steps in both axes. So I wonder if you aren't really at 2x this, but your pixel scale seems correct (based on your pixel size and 571mm focal length)

Second, on your longer runs the values are low but your RA is noticeably larger than your Dec, which would result in star elongation. Not always, but enough to be noticeable

segment 10, 11min

Segment 14 9 min

Segment 21 45 min

So I suggest updating to the latest dev release, and if you didn't create a new profile via the wizard and use the calibration assistant, do those things just to confirm your results. 

Also a few other minor cockpit points:

- fiddling with the values like aggression isn't going to do anything  meaningful, so I would avoid doing that

- it looks like your mount's primary period error is around 667 sec? That period is also the highest residual (although still quite low), so you may benefit from using the PPEC algorithm, set the period to 667 sec and disable auto adjust period.

residual (uncorrected)

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 on the web visit https://groups.google.com/d/msgid/open-phd-guiding/0a85bcc2-02f3-4fa6-833e-a4d67ee8a2f5n%40googlegroups.com.

--

Brian 

Brian Valente

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

nature portfolio brianvalentephotography.com 

other astronomy https://www.brianvalentephotography.com/astrophotography/ 

[Bruce Waddington]

Hi Jim.  Well, I think you were right to be a skeptic here.  I think you've probably got internal reflections in the OAG that are creating pairs of false stars.  Statistically, they look ok but look at their spatial distribution on the guide chip:

MultiStar: List (12):

{X-pos, Y-pos},(SNR)

{310.89, 953.63}(194.8),
{310.87, 413.67}(179.4),

 {1270.86, 413.20}(166.7),
 {1270.92, 953.23}(172.6),

 {525.54, 678.11}(150.3),
 {525.44, 137.97}(170.8),

{1485.58, 677.58}(160.2),
{1485.51, 137.52}(166.1),

{1636.44, 624.01}(128.8),
{1636.42, 84.02}(136.3),

{676.32, 624.37}(133.4),
{676.35, 84.49}(132.1),

You are getting pairs of bright spots that are aligned on the X-axis and spatially separated on the Y-axis.  I don't think we can believe that.

How did you focus the OAG?  Since it's a moony period, you can use the moon to get a rough focus, then move to successively fainter stars to refine the focus.  Also, if you were doing this last night with a bright moon, you might be getting internal reflections that won't show up under darker skies. 

Hope you can get it sorted out - and nice work on realizing it was too good to be true.

Regards,

Bruce

[Brian Valente]

Hi Jim

I assume this was an old response, Bruce gave you excellent advice on the reflections and I see this has progressed

On Tue, Mar 26, 2024 at 8:08 AM Juno16 <jras...@gmail.com> wrote:

Thanks a bunch Brian!

Yes, I did create a new profile inPHD2 for the oag. 

I did create some images and generally the stars look good. However, I can see some elongation in the smaller stars. Not too horribly bad though.

I really appreciate the info on the  primary period error. I did use ppec with the guide camera and had very good results. Thanks for the recommendation. I did want to stay pretty much with the defaults created in the new profile although I did play a bit with the aggression, but found that it did not effect guiding much or at all. Thanks for that info as well.

Were you able to see the four camera star images? Or, was I mistaken? That really threw me because the stars seem to mirror each other in each image and I thought maybe that multi star guiding might have gotten confused by the multiple images and it threw the algorithms off.  

So, in your opinion, the guiding values are reasonable? Or should I say that they look correct? 

With the guide camera, I was getting around 1 arcsec total rms error and with the oag around 0.3. Huge difference!

I posted to this forum awhile back about the high rms that I was seeing and Bruce responded that my seeing was very poor and I might have to lower my expectations. He did recommend off axis guiding.

He also mentioned some stiction in DEC, so I rebuilt that axis with new lube and bearings. I have the bearing set for RA, but haven't installed them yet. I didn't want too many variables at once. Possibly replacing the RA bearing might help reduce the RA error and help with the stars? 

I will also set the max adu as you suggested.

I can't thank you enough! Especially for responding on a Sunday when I know that there are other things that you could be doing!

Much appreciated sir!

Jim

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/7a38b428-35ec-4b1d-9312-5dfafe7ede0bn%40googlegroups.com.

[Juno16]

Hi Brian and Bruce,

Looks like things are working out!

I had the chance to get out again tonight for some testing. I created a new profile (using the wizard as Brian suggested), Changed from the ASI Camera ASCOM camera driver to the ZWO ASI Camera driver and also made sure that the camera was set to 8 bit. 

Calibrated, ran the GA and started guiding on the Leo Triplet just to test things out. Just one panel! No 4 panel display!

Guiding is going very well and much more believable at a total rms error of 0.5-0.6 arcsec/px. 

Many thanks for your help and also to Bruce for recommending going with an oag (in a previous post) because of my very poor lp and seeing.

I am sure that there are some things that I can do better, but this is a wonderful start.

Thank you both!

Jim

[Brian Valente]

glad to hear you are on the road to recovery, as they say ;)

please post a guidelog when you're done, we might be able to offer additional suggestions. 

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/e767bd0a-502d-4b08-b18f-3c95944aad73n%40googlegroups.com.

[Juno16]

Hi Brian,

Here is the log from last night.

https://openphdguiding.org/logs/dl/PHD2_logs_GApq.zip

I really didn't want to take up more of your time. Both you and Bruce have already spent quite a bit on me!

I also changed the RA algorithm to PPEC as you suggested.

My apologies for the choppy log. I didn't have time to let things run for a long time. The last session is probably the most useful as I guided uninterrupted for almost 30 minutes.

I did have a challenging time getting a decent calibration. Still don't think that I accomplished that very well, but enough so as to test things out. I was so happy to see only one set of stars!

I do have more RA error than DEC, but not any worse than I had with the guide scope. I was guiding just north of the CE which is an area that I have really been struggling recently.

Some subs have stars that aren't perfectly round when zoomed way in and I tested eccentricity in Pixinsight. I compared the eccentricity values to previous subs that I imaged with my guide scope setup and they are very similar.

WBPP seems to really clean that up significantly.

I do still plan to rebuild the RA axis (as I did DEC a few months ago) and hopefully that will help some.

Thanks again to you both! 

I am sure that there are things that can be done better and I appreciate the offer to look at the log. I do have something that I can work with now.

Jim

[Brian Valente]

Hi Jim

>>>>I also changed the RA algorithm to PPEC as you suggested.

yes but you needed to set the specific period of 667 seconds and turn auto adjust period to OFF. right now it's at 1106 seconds which isn't really helping you

Once you sort out RA more, you should also enable Dec backlash compensation and set the value to something like 850ms to start (2500ms max). Your mount exhibits a fair amount of backlash, but there's not enough compensation in there to take up the slack when reversing, so it's taking too long. I would not do this now as your RA is already higher, so it would increase star elongation. See if the PPEC will help

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/9b57a38a-3058-41f8-ba96-2ab2fa1c0670n%40googlegroups.com.

[Juno16]

Hi Brian,

I absolutely did forget to enter the PPEC values that you provided me with. Actually, my only intention last night was to only set the mount up facing north and watch the PHD2 window and hope that I didn't see four copies of stars!

I was thrilled that I only saw one set of stars and went ahead and PA'ed and started calibration and running the GA. It was all gravy to me at that point!

I backed off the DEC backlash (adjusted the worm carrier) because Bruce mentioned in a previous post that I might have gotten too aggressive with adjustments (trying to get it as low as possible) and might be causing a bit of stiction in DEC. 

I probably backed the worm carrier a bit too much, but to me it seems that PHD2 was bringing DEC back to the baseline reasonably quickly.

I just entered the values that you gave me for PPEC and DEC backlash and exported the profile for safe keeping.

I can't thank you enough for all of your help! Take care and have a find day!

Jim

[mj.w...@gmail.com]

Hi Jim

I'm not sure I've read how you fixed the 4-pane problem ?

Bruce thought internal reflections.

But the symmetry and similar brightness causes me suggest a camera processing fault.

Although I've seen the split frame situation, where left and right halves of the image are swapped.

I've never seen that 4-pane problem before.

Michael

Wiltshire UK

[Juno16]

Hi Michael,

I was completely at a loss. I posted the issue on both Cloudy Nigh Forum and the Astrobin Forum.

No responses on the AB post, but I did get an interesting response from a member on CN that said that he had issues (didn’t specify what it was) with using the ASI Camera (ASCOM) driver in PHD2 and recommended using the ZWO ASI Camera driver (native). 

Brian at the PHD2 Guiding Forum suggested creating a new profile (using the wizard) and other settings changes (more related to guiding performance).

I created a new profile and used the ZWO ASI Camera driver. By the way, if I didn't mention earlier, I am using a new ASI220mm mini. I also made sure that I set the bit depth to 8 bits. It is possible that I had the original profile (the one where four "panels" showed) to 16 bit. I cannot check now since I don't have the camera connected.

I employed some of the settings recommendations that Brian suggested relating to guiding performance, but not all. I was primarily interested on camera performance at the time.

That was it. Tested the night before last and everything worked fine. Guiding performance was good (might even be better next time out as I have since implemented more of Brian's suggestions.

Hopefully I will never see the 4 panel issue again!

Thanks,

Jim

[Juno16]

Hi Again Michael,

I have read several posts on the internet relating to using 8 bit vs. 16 bit for this camera in PHD2. That is why I made sure that I was using 8 bit for the new profile.

I connected the camera and checked the original profile that used the ASI Camera (1) ASCOM driver and it is set to 16 bit.

I made sure that when I created the new profile using the ZWO ASI Camera Driver, that it was set to 8 bit.

Could be a fluke, but I hope to get out soon to confirm proper operation. 

If I see a reoccurrence of the 4 panel issue, I will post it here.

Thanks,

Jim