Many thanks, Yes that makes sense as could be as you say it polar aligned at start but then settles out of place, on moving to the target to be imaged. I think I need to get it on more solid foundation as maybe they are sinking a bit due to weight and centre of gravity over time shifting, hence affecting the dec. I will try to record the guiding next time and then can see more info.
Pixinsights seems to be finding lots of stars, but still not aligning them, here is from one sub
Reading FITS image: 16-bit integers, 1 channel(s), 3008x3008 pixels: done
58 FITS keywords extracted
Structure map: done
Detecting stars: done
121 stars found.
Matching stars:
88 putative star pair matches.
Performing RANSAC: done
77 star pair matches in 38 RANSAC iterations.