Hi,

Last night I tried guiding for the first time. I was surprised to see the internal guider (KStars 3.6.0 on Sellarmate 1.7.2 x64) push the guide RMS starting with 2 arcseconds to as much as 10 arcseconds eventually. From the graphs I've seen many show sub arc-second RMS!

So, I then tried with PHD2 ... and while there was an excursion to an RMS of 2" at some point, most of the night it was around 1.25" and often closer to 1".

What's going on? I tried with and without GPG, everything else was left at default. I thought there was something wrong with the mount!

Details of the setup:

• CEM26 driving about 18lbs of gear (SV503-102mm, Canon D7, SV106 50mm with SV905C) ... the load is NOT perfectly balanced, I need to do 3 axis balancing and it's ... not trivial.
• Ekos guider is set up to pulse-guide the CEM26 ... and I just realized I have the ST4 plugged into the 905C and CEM26 ... I don't get a clear night for a week, so I'll try disconnecting this and see if that was causing a problem (although I switched to guiding through the SV905C as well and that didn't fare better)
• I have also noticed the image in PHD2 looks different from the one in Ekos; Ekos doesn't allow me to adjust the gain directly in the guider, and neither does PHD2. I don't know what settings PHD2 used for the SV905C.
• Ekos was connected to the SV905C through INDI, while PHD2 refused to do this and I connected it directly to the SV905C.
• Both applications spoke to the CEM26 via INDI.

Unfortunately in my frustration I did not save any guiding images to share.
I would prefer to use the internal guider because everything ties in nicely with the scheduler ... perhaps Ekos and PHD2 know how to talk to each other, but I haven't discovered any way to make that happen.
Also, PHD2 seemed to do a lot more calibration than the internal guider.

So Just to recap. With internal guide you have a bad final result than with phd???

Yes, with PHD2 I get 1.2" RMS, with internal guider I get anywhere from 3 to 10.

"perhaps Ekos and PHD2 know how to talk to each other, but I haven't discovered any way to make that happen."

Ekos and PHD2 can talk to each other. You just need to tell Ekos you're using PHD2 in your profile. Open the profile editor and change it from Internal to PHD2.



"I have also noticed the image in PHD2 looks different from the one in Ekos"

In PHD2, the image brightness is controlled by the gamma slider below the image. It doesn't change the gain of the camera, it just changes how it's displayed on screen. That's probably why the images look different in Ekos and PHD2, even with the same gain and exposure duration.

In PHD2, I believe you can change the gain of the camera by going to the Guide menu at the top of the window, and selecting Camera Settings.

Can you upload your guide log from the guiding run using the internal guider? The logs are stored in your home folder, in ~/.local/share/kstars/guidelogs. Maybe we can figure out why the internal guider isn't working well for you.

-- Kevin

I experienced the same guiding problems with 3.6 but it seems to have been fixed in 3.6.1 beta. I tested it last night and was back to my usual 0.4 arc seconds. Version 3.6 had shot it up to 2 to 5 arc seconds! Glad it’s fixed now.

I go back ti the 3.5.8. Kstars 3.6.0 in my opinioni is not mature to be stable

Thank you everyone.

Kevin, the guide logs are attached. I'm looking at them using the PHD2 log viewer, though I don't understand much from there for now

FWIW, I'm not aware of any fixes to internal guiding between 3.6.0 and the current beta that might improve the default performance.
I did make a change to to allow more SEP MultiStar reference starts, but you have to change that value yourself (it defaults to the old value of 10)
and I made some minor changes to the layout of the guider tab, but other than that I don't know of anything that should have improved guiding in 3.6.1 beta.

I also have not had any issues myself with guiding in 3.6.0 (nor 3.6.1). I have seen a few reports of issues on the forum,
I believe you folks, but I'm not sure what to say about it. (I only use SEP MultiStar).

Perhaps the issues were with some other part of Ekos or INDI that manifested itself in guiding RMS?
Also, were the captured images themselves worse, or just the reported RMS?

Hy

The RMS was worse, the captures were simply different from PHD2. I wish I could retest but I won't have a clear night for a while. Let me know if you want me to turn on any sort of debugging/logging and once able I'll collect some data.

That is strange then. I can run 3.6 and get horrible guiding (unusable) and then quite and start 3.5.8 and leave all settings the same and the guiding RMS drops from 2 to 5 to 0.4 - which is what I am used to with this CEM120 EC2 mount. But as I mentioned before, my starts were round in both versions. Only the RMS chart seemed to be messed up. the chart looked bad - but the reality was the guiding was good. I was able to get some imaging done by changing the guide deviation from 0.8 to 3 arc seconds and my images were lovely. But my images will show errors above 0.8 (I am imaging at 0.4" per pixel) and even when the guiding was showing 2 to 4 arc seconds of error, my starts were sharp. Strange.

I had the same problem with 3.6 nights until Aug 27th version. (Build: 2022-08-27T18:26:30Z)
PS: I am using the Mac OSX versions only.

Right now as I type this, using the AUG27th version of Mac OS X KStars I am guiding at 0.36 and 0.28. I am seeing limited tonight (I live at 7,000 ft) with 15% humidity.

George,

I looked at your guide log. I can say that the calibration runs looks horrendous, and inconsistent from one calibration to the next.

For example, one of your calibrations



Just for comparison, here's my most recent calibration



Also, I would enable the option to re-use calibrations, and then calibrate near the celestial equator (Declination 0) and the meridian. That's where you'll get the best calibrations.

I'll try that...I am able to go as low as 25 degrees or so.

I see what you mean about the calibration. For whatever reason, PHD2 did better. My telescope wasn't well balanced either, I have just worked out the complete 3D balance for it. I'm going to report back once I have a change to try this again.