This is just an idea of a beginner, I don't know if it is viable.
Several mount has a PPEC that has to be trained while autoguiding for a worm period. However it would be better to train the PPEC using an average of several periods and a smoothed curve. This is where PemPro comes useful. Since GPG RA guiding, to my understanding, builds up some sort of fitting of the mount periodic behavior over several period, I was wondering if it would be possible to use this data (maybe with some processing to cleanup and detrend the curve) to obtain a better mount PEC training.
Any feedback on this? I have a GEM45 and right now the only option to program PEC is during actual guiding, with the result that a noisy and maybe drifting correction is recorded. Since ioptron mounts does not allow PEC recording from PC also using PEMpro is problematic. Using a trained GPG model to program PEC during guiding with a cleaned-up curve would be a nice and unique feature of Ekos.
I have modified the iEQ driver to allow for SetPEC on and Off as well as train.
My thought was exactly the same as yours, for now you can guide for 5 or 6 full worm cycles on GPG or PHD2 Predictive.
The GEM45 worm cycle is 400 sec, so you would need to guide for about 2400 sec (40min) then start your record for a smooth predictive curve.
Once complete, stop the guiding and restart GPG.
This means that you don't need to park and can setup and go with PPEC on next time.
If you want to test the driver ping me.
Christopher | Sussex England | n2042.blogspot.com
iOptron GEM45 | Altair Mobile Pier | TSPhoton 154 @ F3.7
ASI 294MM Pro | ASI EFW | ASI EAF | ASI 120MM
Mini PC | Quad 4GB 64GB 5G | Juice USB Harness
Interesting! Do you have your own repo with the modified driver?
Actually I was thinking something more. When you guide you are correcting both for PE and drift. Ideally you should program only the correction for PE, neglecting drift, which may depend on other factors. I experimented a bit with PHD2 simulator, and tried to set PPEC to 100% and reactive to 0%, but apparently in this way the PE is not completely removed. I had no opportunity to test on field. Do you know if with Ekos it is possible to have a 100% predictive guide? It seems that it is not possible to simulate PE error in Ekos (asked some time ago on the forum), so I could not try.
Christopher | Sussex England | n2042.blogspot.com
iOptron GEM45 | Altair Mobile Pier | TSPhoton 154 @ F3.7
ASI 294MM Pro | ASI EFW | ASI EAF | ASI 120MM
Mini PC | Quad 4GB 64GB 5G | Juice USB Harness