I believe - ASI120 VIA ASI120 means ST4 port on the camera. and ASI120 via [your mount] is internal pulse guiding. I use the later.
PHD always liked a weak star. If I pick a weak star in the internal guider - it drops my selection and picks a BIG BRIGHT ONE?
The internal guide program likes big bright ones. PHD wants a weak start that has a nice pointed PEAK at the top and not a over exposed flat top.
I have good results with PHD (on windows my old setup) I tried PHD on the PI - as a server to the desktop (Kstars/ekos) in the house. It worked, but
not nearly as good as being on the local machine. I suspect it has to do with the PI 100 yards away running PHD and then sending the results to me
on the desktop in the house. IT FELT... LAGGY and guiding from (local at about .35 dropped on the PI to about .79. The internal guide code also
guides about .79. so I think it has to do with the REMOTE server delays.
I had problems too - but they just went away. We seemed to think it was due to lowering from 10 iterations to 3. It also complained I have backlash
but I have a high end mount and really doubt that.
The GUIDING code is the weakest link in the Kstars/Ekos software pack. Hopefully it gets a refit. PHD is open source why not combine it?
PHD has been combined into Asiair and works extraordinarily well with my cheap AVX mount I get around .5 total rms error. (with sighting down the tube to PA).
I'll attach - I have a PHD guiding on windows directly and I think I might have a cap of guiding with the internal Eko's guider.
PHD - typical direct from NUC PC in dome to mount.
Internal Kstars/Ekos guiding. Best I've achieved so far.