I’ve been having a lot of problems lately when changing from one object to another in the sequence. Every time I finish a sequence and do platesolving to change to another object, the program doesn’t find a star and throws an error. The star is perfectly fine in the PHD2 but for whatever reason the image doesn’t refresh and I get a lost star error. As soon as I detect the problem, touching something in the PHD2 like reactivating the camera, for example, seems to reset the SPG and start the session recovery routine. Then it waits a while, does platesolving again and this time, the second time, it does find the star. What could be happening?
This also something I have been experiencing frequently the last couple of months. Trouble is that it is an intermittent problem, so somewhat difficult to analyze. I can see PHD2 pick up a correct guidestar (good profile) and then immediately lose it. If I am quick enough to push autoselect star PHD2 recovers, but then SGP already goes into recovery mode. Trying again in SGP results in another (unnecessary) slew and center and then the same issue with PHD2. Only way out is to abort and restart the sequence, but then without slew and center on the first (current) object.
I’m afraid there is no easy answer here, but if it’s never happened before is there anything at all in your rig / system that has changed recently?
Anything different with your PHD2 profile?
ASCOM platform upgrade?
Drivers?
etc
Is the issue repeatable in different areas of the sky?
From the logs, I can see that PHD2 is successfully restarts and then, as you say, immediately loses the star. Have you been able to observe this in the PHD2 app itself? It may offer clues.
In terms of possible SGPro bugs:
I notice that you are not using dither or settling. This is totally fine, but atypical. It has me wondering if the absence of those options introduces some kind of timing issue, but, for an issue like this I would expect to see SGPro fail and all I see is an error directly from PHD2. In any case, you might see if you observe different results when you enable settling (don’t worry about enabling dither). I am not implying that, if this works, that it’s the solution… we’re just gathering evidence at the moment. Has it been on in the past?
In general, and I don’t know if any of this applies to you:
Does PHD2 connect to an ASCOM driver for its mount connection? There are non-ASCOM connections available, but ASCOM is, by far, the most reliable and resilient connection type.
When PHD2 is asked to auto select a star does it lock on to a real star?
Have you made sure that the PHD2 exposure length is sufficient and that the local stretch function produces clean data?
How are you calibrating? If re-using an existing calibration, maybe delete it and re-calibrate from scratch.
Anything that may be preventing PHD2 from pulse guiding.
my old sgpro is also happened this problem. someone hs found the solution. They set the recovery mode on. When that is lost star, it recover the sgpro, and then it is work. It seems to be not prefect solution, but it is work.
Someone say that is caused by the PHD2 and SGpro timelapse issue. that is SGpro rare bug. Sometime that will be happen, not not always. case by case.
Thank you, Ken, for following this issue. Yes, Ican give you more information about this problem. I will try to follow your questions:
Dithering option: I usually select this option but for the project I currently involved I have had to remove it from my set up. However, I expect not to have relationship between dithering option and the lost star event even I might think having dithering it is worse for this purpose, even. I will check it.
I use ASCOM Drivers last release, I guess. I’ll check it.
Exposure is largely enough with SNR > 100 in most cases.
I can do more calibrations but as previuos reasons I don’t find relationship between events. Never had issues in the past.
For further information I’m trying to exactly describe the process because I had the chance to see with SGP and PHD2 opened windows at same time:
Please, imagine I’m taking the last frame of a secuence. Everithing works fine, PHD2 guiding and the guiding star (with a squrare with a big green cross in the image) an the others stars (with small circles) are identified.
Let’s call the green references in PHD2 of the guiding of this object as “green references 1”.
Secuences ends, then image in PHD2 freezes: stars and green references 1.
SGP performs a “center object” process for next object.
Center object performed.
SGP call PHD2 to resume guiding
In PHD2 I was able to see how the camera switched on again, the frame was updated but not the “green references”, that is, the previous “green references 1” were kept for unknown reason and the PHD2 was trying to find the star exactly in the same position the last object had. There fore, due to the very low likelyhood of having a star in the same position, the result is a “lost star” event.
After that the “recovery process” usually starts. After a minute SGP tries another “center object” , resume the guinding and in this second round ususally I can see how both, star image and green references, udate in phd2 and things go right again.
This is why I was wondering if having the “settling” option disabled in your sequence may be causing a timing issue. When settling is on the sequence will wait until the guider reports it is ready and then resume the sequence. When it is off, the sequence will resume immediately and the guider may not be ready (and then the sequence may fail)
Well, probably I didn’t read all your advices correctly. My apologies. Which setting option are you referring to, exactly? I’ll try next clear skies night. Thank you in advance.
Hi, Ken. Yesterday I had a chance for testing your piece of advice. I selected to settle for 3 seconds after dithering (with dither unselected). I think you were right. It seems the problem is solved. I didn’t have any issues with finding a star in PHD2. Cross fingers to continue in this way.