Seeing Alfred's guiding makes me wonder what the problem could be, I thought it was my mount but couldn't find anything wrong..
This is from another session..

Markku,

Yep... that looks pretty familiar. Looks like some kind of bug in computing of the pulse.

I'll continue to use PHD2 until fixed.

Thanks,

Ron

I can see where the huge guide pulses are sent. This is the guide log. Do you have the Ekos log, too?

No, I forgot to turn on log to file that night.

If it's helpful, I can grab the EKOS/INDI logs that match the original post guide log tomorrow.


Ron

Ron,

your graph shows constant over correction starting 20:44. Normally this would happen if aggressiveness was set too high or the guiding rate was wrong. But in your case it happens suddenly after everything was running smoothly for quite some time. You certainly won't change your aggressiveness or guiding rate settings while guiding so I have no idea what's going on.

Alfred,

Yeah, it's almost like the guiding program multiplied the pulse needed all of the sudden.

Still trying to find an answer.... but thanks!

Ron

Ron,

next time it happens I would try a few things:

1. lower the aggressiveness in RA to like 10% and watch whether or not it has any effect on guiding. At the very least it should lower the amplitude.
2. If you haven't done so already, switch off all 3 "predictive" methods, i.e. Integral Gain (set it to 0), GPG (uncheck the box) and "intra frame dark guiding". (I think the latter is deactivated when GPG is deactivated)
3. In Indi check your Telescope/Guiding Rate setting. Just to be 100% sure. It should match the setting in your mount/handbox.

I admit chances are slim that any of these measures will correct the problem. But step by step we should hunt this bug down.

Alfred,

Thanks for the suggestions. I have a couple of questions/observations...

1. While it was in the "death spiral", I tried changing the aggressiveness setting. It didn't seem to make any difference. It was like it was completely off the rails and computing a pulse that had nothing to do with guiding error or settings.

2. I haven't tried this... GPG was checked for the entire time. Hmmm... you thinking that the big movement is an error with the prediction factor vs the current guiding error? Yeah, I'll try this.

3. Yep. I don't use a handbox, but I double verified the settings.

Again, the weird thing is that it can work fine for hours... then boom! It seems to have only gone crazy just after dithering. Hmmm...

Thanks,

Ron

Ron,

1. excellent point! And exactly right, the aggressivenes setting has no effect on the predictive component. The fact that altering the aggressiveness had no visible effect on your "guiding in death spiral" suggests to me the predictive component COULD be the problem (I don't yet think it IS). So let's switch the predictive component off completely and see what happens.

I wish I could run tests here, too. Unfortunately weather doesn't allow.

Alfred,

Okay... finally got a clear night to test some things. I ran with the internal guider and got the same results fairly quickly. I have attached the EKOS log and guide log for your examination. The problem appears at 19:45:26.913, where the RA and DEC pulses become really large and alternate in direction with every pulse. I tried changing settings while it was in the "death spiral"... turned off GPG, stopped predictive guiding, min move, etc. Nothing stopped the craziness. Finally, I changed the algorithm from multi-star from SEP Multi-Star to SEP. That stopped it.

So,.. when I look at the log just before the craziness, I see a line that says "MultiStar: Failed, fell back to guide star". I don't know if that is the cause, but since it appears just before the large pulses... makes me wonder.

Hopefully, someone that handles guiding can take a look at this.

Interestingly, when I restarted SEP Multi-Star guiding it worked fine for the next hour. Weird.

Thanks,

Ron

Ron,

I had a look at your logfile and haven't come to a conclusion. What strikes me is the fact that the guide star is lost although SEP delivers 100 stars as usual so star detection is not the problem. I'll try again later tonight.

Thanks!

It's unfortunate that I'm having this problem. The internal guider does a splendid job of guiding (better than PHD2), but with this issue I can't use it. It's also weird that it is so random.

Ron