Hmm! I wonder if this is the cause of my recent guiding issue.
On consesutive nights I ran the same rig with the same guiding settings, but the guiding on the 2nd night had PHD2 changing from ~2 sec to <0.5 sec updates with a 2000ms setting. Changes to the refresh time (now set at 5000ms) made no difference. Over the 5 hours, though, there was a steady drift of the field to the point where my stacking routine just gave up, and PHD2 didn't report anything unusual!
I'm wondering whether I had got my guiding working at a point much closer to the equator on the first night, whereas I did it near my target at ~-70deg. on the 2nd night. As this was about a month ago, I can't remember that sort of detail. I'll have to try it out and let you know.
Paul.
PS Apologies in advance: this is a bit off your original topic, but might be of interest to others. Let me know if I should start up another topic.
Update: No, this isn't the same issue. This is simple user error (I say with a red face!). In the PHD2 calibration screen, I (somehow) had selected the 0.5sec exposure time. I've now put it at 3sec and all is working well. A 6-hour ap session last night couldn't have gone better! I did, however, calibrate my PHD2 while pointing at Achernar (Dec -57d), rather than the -70d of NGC2070 (Tarantula Neb.). Total RMS error for the night seems to have been mostly between 0.5" and 0.8".