Images start to drift in RA at end of long multi-filter sequence

This is an odd problem that had occurred about a month ago, and I just thought it was a fluke, but then occurred again last night. I’m using SGPpro 3.2.0.613 with a EdgeHD11 on a AP-1100 GTO mount. Guiding via PHD and OAG. I had set a sequence to capture M90 with 20 RGB subs followed by 40 Lum. Weather was not great, there were some clouds which did require SGP to go into recovery mode a few times, but when the problem occurred the skies appeared to be clear.

All the RGB subs showed great guiding - typically no more than 0.5 to 1.0 arc-sec (not bad for the sky conditions. Then, when the Lum subs started each image started drifting in RA about 1.8 arc-sec per sub (120sec exposure). DEC was spot on.

The Lum Sequence started at 00:45 on 5/14/2021 and ended at 03:02. (Complete sequence started at 22:09 on 5/13 and ended at 04:29 on the 14th.

Guide cam ZWO174mm; AP software APCC using the APPM Model for tracking and guiding corrections; Stellarium and Stellarium Scope for pointing.

I’m at a loss to figure this out, as it could be SGP, PHD or even the AP software, though that is unlikely as I’ve demonstrated almost 2 min subs unguided (but you never know).

I’ve pasted links to both the SPG and PHD logs as well as the seq file at the end of this message. I will also post this to the PHD forum as well.

Can you expand on what you mean here:

  • How are you determining / measuring the RA drift?
  • Are you seeing trailing in RA in the captured images?

Looking at your PHD2 logs, there is no visible change in guiding behavior when you start capturing LUM. You appear to get about 1 hours worth of imaging before clouds come in and you get a sequence of lost guide stars.

And your SGP logs show the same: about an hour of imaging from 00:45 through 01:45 before guide stars are lost.

If you are seeing trailing in your final images but guiding continues to look good, then that would normally indicate differential flexure. But you’re using an OAG so that seems unlikely. Is it possible there could be cable drag pulling the guide cam…?

After examining the PHD logs with the folks at PHD it seems that I had the Minimum star HFD (pixels) set to 0. Hot pixels and the like were probably selected during the times the image drifted. I have re-examined all my PHD parameters and recent sessions have been successful.

It might also be possible that drift occurs if you pause autoguiding during centering or focusing, and then PHD2 selects a different star. Some other apps have an instruction to plate solve an exposure and if its center has drifted too far from the original target position, it does a centering action before the next exposure.