Well at least I'm not alone with this irritating issue. But joy, I may have found the problem.
I've been experimenting and I think I can confirm the following:
- It is specific to the RA axis (reproduced the effect with slewing and turning off tracking).
- It does it in both directions (it isn't just the axis stalling, it will sometimes go the other way) - important observation.
The attached chart shows five of these "events". On the first three it completely lost the star and I chose another one (so while it looks like it got the star back, it didn't). Note the "effect" goes in both directions. On the forth, PHD fought and managed to get back tracking. I stopped on the fifth and last event. The mount was pointed for drift aligning (pointed at the zenith? and meridian I guess).
What this suggests to me is that it is not a stall. It doesn't sound like one and a stall couldn't actually actively slew the axis, which seems to be happening sometimes.
This leaves software or electrical problems. I've rebooted and still got one of the events. I also looked at the Eqmod log and saw nothing being sent except the correction from PHD.
But I also finally rerouted the shoestring cable thing away from everything else electrical and gave it its own USB port (not through the hub). I'm currently looking at the computer in the observatory (Teamviewer is quite good...
) and its been running for nearly 20 minutes and there hasn't been one event!
So it seems to be electrical interference on the shoestring cable.
Of course a bit early to call it definitively, but definitely promising.