PHD2 does not auto-find guide star

[Jerry Hulm]

I recently upgraded PHD2 from 2.6.10 to 2.6.11 and now PHD2 keeps reporting it cannot auto find a guide star.  Also, I have recently switched to using NINA (used SGP previously).  I suspected there was a problem with how I configured NINA to work with PHD2 but the interface is too simple.  I also have new SX cameras (main and guide – the guider is a SX Lodestar Pro), but have guided successfully for the past couple of months with these cameras with no issues.  I did adjust the parameters in PHD2 for the new camera and my imaging train configuration – using an OAG and a 2x barlow.

When I start a sequence, I can see that NINA is starting up PHD2, PHD2 is looping but then reports that it cannot find a guide star.  I can click on a star and PHD2 will guide fine. But when NINA stops PHD2 guiding during an autofocus, PHD2 fails (and fails and fails) to auto-find a guide star, and I have to manually select a star in order to get things going again.

The HFD of the star I manually select is well above the minimum I had set in PHD2. And from what I can tell looking at the log file, PHD2 doesn't seem to see those larger stars.  I did fiddled with some of the other parameters in PHD2 to try to get things going, so I may have messed something else up in the process (mostly the min HFD size and exposure times).

I have a Lodestar Pro (pixel size 8.6 x 8.3uM) and use their ASCOM driver – I am using bin 1 for my autoguider. I did load 2.6.11 Dev1 but have not had a chance to try it out yet.  I should also go back and try the same session on SGP … but skies have not cooperated recently.

Attached is my log and debug files from the failed guiding sessions.

Thanks for the help.

Jerry

[Bruce Waddington]

Thanks for opening the new thread.  I don't think the latest release changes have anything to do with this problem but I think we can get you sorted out again. The lost-star events aren't triggered by the Min-HFD parameter - they're coming from star-mass change events and from stars that don't meet the minimum brightness threshold.   I don't understand what equipment changes you made - are you saying you're using a 2x barlow lens with an OAG?  If so, that's probably the source of the problem.  Your log shows that the stars seen by PHD2 were really faint and weak, peak ADU levels were typically below 1700 adu - and that's with a 16-bit camera.  With your focal length scope, there's no reason to use a barlow lens.  The barlow lens is likely to create a lot of light-loss on the OAG and it's unnecessary for guiding.

If you can clarify this business with the barlow lens, I can probably give you a procedure that will get you back to guiding.

Regards

Bruce

[Jerry Hulm]

Bruce,

The barlow is in the main imaging train.  Takes the focal length to 1640 mm for my refractor.  The Lodestar is mounted directly to the OAG.  The prism for the OAG sits in front of the filter wheel (SX Midi Filter wheel).

Jerry

[Bruce Waddington]

Ok, that's even more surprising - don't think I've encountered anyone trying to image with a barlow lens except for lucky imaging.  If that's a recent configuration change, it's still likely to be the source of the problem because it reduces the field of view by 2x and drops the apparent star brightness proportionately.  But no matter, here's what you can do to recover the guiding:

1.  Upgrade to the 2.6.11dev1 release

2. Make a new PHD2 profile using the native SXV camera driver and bin the camera 2x2.

3.  Create a new dark library and consider creating a bad pixel map as well

4.  Disable star-mass detection in PHD2 (Advanced settings/Guide tab)

5.  Set the Min-HFD parameter to 1.0 (Advanced settings/Guide tab)

6.  Make sure the guide camera is critically focused - you'll need to use some kind of focus assistance for doing this, just eyeballing it won't be sufficient.

7.  Test auto star selection and look at the appearance of the guide camera frames under reasonable conditions - not near a full moon for example.  You can capture a typical guide camera frame using the File/Save menu item, then examine it in whatever image viewing tool you prefer.  This testing should be done manually without NINA or SGP banging away trying to image - they aren't causing the problem, they just complicate the trouble-shooting.

If you still have problems, send me the new  guide and debug log files (Help/Upload Logs...) and I'll look at them again.

Regards,

Bruce

[Jerry Hulm]

Bruce,

Made the changes you suggested and spend Tuesday evening cleaning up my focusing and correcting some tilt issues.  Once I completed all of that I started an imaging session and everything seemed to be working fine.  I tried another imaging session Wednesday evening on the same target and had the same issue recur with auto-star find failing.  I noticed the stars in the guider field looked out of focus, which didn't make sense as I spent a lot of time focusing the guide camera the night before.

I do notice that when I start an imaging session I always have to move the telescope focuser about 100-150 steps in order to get it back in focus from the previous session (my focuser is a Microtouch - there are about 300 steps per millimeter of focuser movement).  I have always attributed the change from the previous evenings focus position to a warm vs. cold telescope/optics.  So I  was wondering if the focuser movement is causing blurring of my guider images to such an extent that PHD2 auto-select fails.

I will also try to improve the focus the the guide camera again.  Could be I just did a poor job.

Anyway, I attached the logs from Wednesday evening when auto-star find failed.  On the Wednesday session, when auto-star find failed, I did manually select a star when I noticed the auto-select fail and then PHD2 started tracking.  On the second time around after guiding was halted for an auto-focus, I just let things run to see if PHD2 ever selected a star on guiding restart.  It didn't.  Then I shutdown the session.

If you see anything suspect, or any errors I made, please let me know.

Appreciate your help.

Jerry

[Bruce Waddington]

Hi Jerry.  It's a little hard for me to see what's going on but the image quality still seems to be quite poor with only low-SNR stars available. You didn't send the previous night's logs when star-finding was better so I don't have a comparison available.  I still have serious doubts about trying to image and guide with a barlow lens - what is the native f-ratio of your refractor?  I think we need to look at some of your guide frames.  Fortunately, you already have them.  Each time there is an auto-find failure, PHD2 automatically saves the images in separate folders that are stored in the same directory as the other log files.  You should look at these images with an image processing app - they are just fits files - so you can see what PHD2 sees.  I suspect they will look quite poor.  The folder names  include a timestamp for when the failure occurred, so you should have a number of them from the 3/18 guiding session.  If you want me to look at them, you should post them to a publicly accessible cloud storage of your choice.  In the case where you manually selected a guide star, that star had been rejected by PHD2 because it was too close to another, somewhat fainter star inside the tracking region. The real problem is that there was only the single star in the entire field that was clear enough to be a good guiding candidate.  One other thing I will mention is that NINA is imposing a subframe restriction for the auto-find - I think you should disable that, it's unnecessary.  It's not a big deal but with such a small field of view, you need to use the whole field.

Regards,

Bruce

[Jerry Hulm]

Thanks Bruce.

I will do some more work on my setup and look at the data you noted.

I think at this point I know it is my problem ... so will do due diligence in figuring out what I am doing wrong.

I have an f/6 refractor, AP130mm.  I have one of AP's Advanced Barlows.  I really like using it on some of the smaller galaxies and globular clusters.

Jerry

------- Original Message -------

--
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/DhN82BA0A8c/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/483efa5c-6252-4858-ab7f-f1135b3936ffn%40googlegroups.com.

[Bruce Waddington]

I think looking at the images may be instructive.  I would expect the 2x barlow to shrink the exit pupil by 2x so it's possible the pickoff prism in the OAG isn't fully illuminated or may only be sampling areas that are affected by field curvature.

Good luck,

Bruce