Star lost, Guiding achieved only with 0.2-0.5 s exposures, CGX-L problem?

[Alexandros]

Hi all,

Above all I want thank the people who made this amazing software PHD2 (+ helpful documentation) and are so great in helping amateur astronomers. I feel grateful for the PHD2 'fathers' and all other people who offer their knowledge to help the community.

After two decades of dreaming about it, two months ago I could purchase and receive my C14 Edge with the CGX-L. I have started imaging DSOs initially with short exposures up to 35 s. In parallel, I was trying to learn to use PHD2 to make longer captures with the Celestron off-axis guider. I use ASI1600 for imaging and ASI290 for guiding and I am waiting for the ASI174 mini that I have ordered.

THE PROBLEM: I have tried several nights to achieve guiding with little (if any) success. Most of the times I receive the "star lost" message, even if the star is well bright and visible in my screen. Often the calibration cannot be completed. I tried everything I could and read everything I could. two days ago (6 May), I finally succeeded in guiding with PHD2, but this was possible only in a particular area of the sky and the guiding was just above 1 arc" (and with the 0.7 reducer). Even with this "success", my captures were not better (no improvement, and perhaps even worst) than when unguided for captures of 3o seconds only. I had to through away the 2/3. The stars were often elongated always in the same direction...

After reading a lot, I suspected that my mount was the main problem and tried to "train" it and upload 6 runs of PEC to improve guiding. In order to do so, I though that using the ASI290 native (with no reducer; focal length 3910 mm) would have the best possible effect in recording the PE and therefore, I would see the best improvement. With ASI290 I tried 3-4 hours whatever I could, but the star was always lost... I thought using the ASI1600 as guiding camera would be a good experiment. I did so because I observed some rapid movements of the guiding star (although it was never really far away from the red tracking cross) and thought that my mount is not behaving normally. It worked, but only with very fast exposures of 0.5, 0.2 and 0.1 s, which are according to all your advice not a good option. Even so, I did not achieve something better than 1.5 arc".

I am quite desperate and do not know if my mount is problematic or I am the problem and do something wrong. You are my last hope... (before trying to send the mount back for inspection).

I also have a question: When I try to use a bright start (magnitude <9) PHD2 never succeeded not losing it. Neighboring faint stars (magnitude 9-11)  were working much better. Is there a way to use brighter stars as guiding stars? Can we achieve this by reducing somehow the gain?

LOG Files link: https://openphdguiding.org/logs/dl/PHD2_logs_gW9y.zip

Thanks a lot in advance for any guidance or advice you may provide,

Alexandros

PS: I hope I have uploaded the right files... The dates seemed a bit different than when opened with the PHD2 Log Viewer. I have uploaded the log files from the long guiding of 6 May (just above 1 arc", 0.7 reducer and ASI290) and from today when I achieved some guiding with ASI1600, no reducer, some 1.6 arc" (with several interruptions by "star lost"; only one was due to a cloud).

[Bruce Waddington]

Hi Alexandros, thanks for the nice comments.  I don't know if you've done guiding before but if not, I think you're going to be in for a wild ride with this setup.  Imaging with a 3900mm telescope on a mass-produced mount is going to be difficult - do-able but it will probably take some work.  I realize it was your dream system but it's a very difficult way to learn imaging.  Anyway, there are multiple problems here.  To begin, you're trying to guide at an image scale of 0.2 arc-sec/px which is much too small.  When you ran the new-profile-wizard, you would have seen a warning icon about that and a message that you should bin the guide camera.  You evidently didn't do that which was a mistake.  Your guide camera only allows 2x2 binning whereas 4x4 binning would get a better result - but you should at least try 2x2 binning.  Use the new-profile-wizard to create a new profile for binned operations.  The too-small image scale results in huge star disks that are often not recognizable as stars by PHD2.  I also have doubts about how well the guide camera is focused.  Once you make the change, you should spend an extended amount of time just making sure that the focus is precise and you can minimize all these lost-star events.  You must disable star-mass detection in PHD2, that's on the guiding tab in Advanced Settings.  To reach critical focus, you need to use a quantifiable way of judging it - the PHD2 Star Profile tool, a Bahtinov focus mask, or another app like Sharpcap.  This process can take quite a while with an OAG, especially at this focal length, but once you've done it you shouldn't have to repeat it.  The problems you're having with lost stars has nothing to do with PHD2's star selection process and everything to do with your setup.

Next, your calibration points to other problems:

These two lines should be nearly perpendicular, the star tracks should be nearly linear with reasonably even spacing of points, and the overall lengths of the paths should be much shorter.  For some I reason I can't understand, you have chosen to require movement of 100 pixels in each axis rather than the default of 25.  That's a bad idea, you need to leave all these guiding parameters at their default values - when you create the new profile for binning, don't subsequently fool around with any of the guiding parameters.  The clump of star positions at the top (green) shows you have a large amount of backlash in Dec.  The Guiding Assistant should be used to see if you can measure it or at least look at it in more detail.  The curved track of the Dec star positions (green) is caused by a large amount of uncorrected periodic error in RA compounded by the fact that you've forced the calibration to take 4x longer than it should.  If we look at the RA behavior from one of your guiding sessions, it looks like this:

Overall, there is about a 25 arc-sec peak-peak tracking error in the mount.  The bigger problem is the 2 arc-sec error that occurs on a period of 21 seconds.  This will be very hard to guide out and it will put a lower limit on the guiding accuracy you will be able to achieve and is likely to produce elongated stars in your images.  This error is probably coming from one of the spur gears in the RA drive system and probably requires mechanical improvement. 

When you pursue these things, you should focus on just getting a good calibration with no alerts - trying to murder ahead with a bad calibration is just a waste of time.  With all the Dec backlash, you will probably need to clear it manually - with PHD2 looping on a star, use the hand-controller to move the scope north ('up' button) until you see the star clearly moving on the PHD2 display - then start the PHD2 calibration.

Good luck,

Bruce W.

[Alexandros]

Hi Bruce,

Thanks sooooo much for your immediate reply! I really don't understand how can you answer so fast with so detailed answers to the problems of so many people in order to help them. You are an angel! :-) The nice comments are nothing in comparison to the good you are making to the community!

I went through your message and I understood quite a few mistakes I made, but there are some things I still do not understand. Therefore, I come back to you with a few more detailed questions, if this is ok with you. I have been doing only solar and planetary imaging for a while, so the "wild world" of guiding is new to me as you rightly guessed, apart the manual guiding I was making some 30 years ago with my C8 and a piggy-back camera with analogue film in the time of dinosaurs... A Bahtinov mask was in my list; I will try to order one next week.

1) I do understand that the C14 has a very demanding focal length as you wrote. I am not planning to image DSO at native focal length anyway, so would it be ok to try to record the PEC data while guiding through PHD2 with the x0.7 reducer in my train?

2) In the new-profile-wizard I enter the 3910x0.7=2737 focal length. Is this ok, or should I change this value depending on the distance of the camera from the back side of my optical tube? I try to respect the 147 mm back-focus (146 mm + 1 mm for filter use), so if I am exactly at that distance, 2737 is the right value?

3) I am waiting for the ASI174 mini guiding camera in the very next days. Its pixel size is 5.86µ. Should I use 2x2 binning with this camera as well? I made a test in the wizard and saw that with no binning I still get the warning symbol. When I put binning 2, the pixel scale becomes 0.88 and the warning symbol does not show up anymore.

4) " I can't understand, you have chosen to require movement of 100 pixels in each axis rather than the default of 25".  Hmmm... I think I did not touch this. I see it is in the advanced settings (brain) and the "advanced" button of the calibration tab. May I have changed it indirectly through any other parameter? Anyway, now that I had your advice I won't fool around with guiding parameters as clearly written in your guidelines... I did that only as a last 'desperate man's chance' before asking your help. :-)

5) "With all the Dec backlash, you will probably need to clear it manually - with PHD2 looping on a star, use the hand-controller to move the scope north ('up' button) until you see the star clearly moving on the PHD2 display - then start the PHD2 calibration". Can you please explain this a bit more? I don't understand what exactly I should do manually and how this relates to clearing the Dec backlash... Why only with the 'up' button? Place a guiding star exactly at the center of the looping screen of PHD2 before calibration? Is that all, or something more complicate that I do not understand?

6) Is there a particular direction/section of the sky in which I should peak a star to make the calibration of a new profile?

Thank you very very much once more for all your precious help!

Best wishes,

Alexandros

[Bruce Waddington]

Hi Alexandros.  Here are some answers to your numbered questions:

1.  Yes, working with the reducer would be fine, just reduce the PHD2 focal length setting accordingly - but always in a new profile.

2.  Starting with the nominal 2737 focal length should be fine.  If you plate-solve one of your images, you can find the effective focal length and then change the PHD2 value if you want.  It only affects the guiding performance results that are reported, not the actual guiding.

3.  You should bin whatever camera you're using until you can get above the 0.5 arc-sec/px threshold.  In other words, don't trigger the warning in the wizard.

4.  I've looked at the code again and the calibration distance value can be changed if you do anything else that affects focal length, camera pixel size, or binning.  So the insanely low guider image scale you were trying to use probably resulted in the large calibration distance.  The default PHD2 target is to achieve movement of roughly 20 arc-sec on each axis during the calibration.  If you get your image scale up above 0.5 arc-sec/px, things should be ok.  It's equally important that you avoid trying to adjust any of these basic configuration parameters manually, you should always use the new-profile-wizard. 

5. I'll assume you understand the basic idea of backlash and the fact that the gears on your Dec drive system aren't fully engaged in both directions at any given time.  The PHD2 calibration measurement only considers the north moves for computing the Dec guiding variables so we want the north movements to be "clean" - free from backlash and any static resistance.  The calibration process includes some backlash clearing but with a mount like yours, it may not be sufficient.  If you have forcibly moved the mount north and you can see the stars moving on the display, the Dec gears should be engaged in the north direction.  We're not talking about the "sky" sense of north, only the direction the motor runs when you press the 'up' button on the hand-controller. 

6. Ideally, you should calibrate somewhere near the intersection of the celestial equator and the celestial meridian.  That's near the intersection of Dec=0 and the imaginary line passing through the celestial pole, the zenith, and the horizon.  It looked like your calibrations were being done in that area.  If you fix your image scale problem and get the camera properly focused, you should let PHD2 auto-select a guide star.  Forget about bright stars and stars whose name you know - those will be saturated.  Working with faint stars is fine, it's all in the math and PHD2 knows how to select stars that are likely to work best.

Cheers,

Bruce

[Alexandros]

Dear Bruce,

Thank you sooooo much!

Thanks to your precious advice, tonight I could achieve several times to guide without loosing the star. I tried three sessions. The first with the ASI1600 at native 3910 (just as an experiment, since this camera allows for much more binning), then the same train but after resetting to factory settings my mount. In both cases I could upload PE corrections. Finally I tried the ASI290 with the 0.7 reducer at 2737 mm. This was the most successful. I tried many calibrations (many came out with warnings and were really bad; the mount was delaying a lot to make the north movement...) of which the last one seemed decent to me. then the guiding (during a new PEC) was for the first time below 1 arcsec! :-)

I am still far away from a good guiding, but this was a progress. In the next days I will make new tests with the ASI174 min guiding camera when it arrives.

Anyway, I am afraid, that if I do not hypertune (or repair) the CGX-L, I will not achieve really good results. And since it is under warranty, before calling the vendor and take decisions about how we will proceed (I might send it to him or open it myself and try to tune it...), I would like to kindly ask you to have a look at these last logs of tonight:

https://openphdguiding.org/logs/dl/PHD2_logs_Z2Jm.zip

and if possible let me know if you have any additional comments regarding the problems of my mount that I should mention to the vendor.

Anyway, I wish to thank you once more, because you made me smile again with your help and the improvement I could make. At least I can have PHD2 happy without loosing its star now :-)

Cheers,

Alexandros

[Bruce Waddington]

I don't see any new information that relates to the mount.  When you have a calibration that takes a long time to start the north moves, it's because you haven't cleared the backlash sufficiently.  With your mount, that has to be done every time you do a calibration.  Also, I saw that you ran the Guiding Assistant but then didn't apply any of its recommendations for setting MinMoves.  Why not?  And why were you pointing clear down near the horizon when you did all this?  A lot of these early guiding sequences were done pointing down in the weeds below 30 degrees altitude.  When you ran the GA, did you let it try to measure your backlash?  The two big mount problems you have are the periodic error I told you about before - especially the 21-sec gear error - and a large amount of Dec backlash.  Beyond that, there are still a lot of operational errors that are limiting your results to some extent.  The final guiding sessions of around 1 arc-sec total RMS are likely to be good enough for you to learn how to image using multi-minute camera exposures. 

Good luck,

Bruce