Have you considered differential flexure between guiding train, and imaging train? You don't need much to cause problems.
Is the egg shape only in-line with RA direction, or does it vary depending on orientation of the cameras, and the pointing direction?
Is it worse or better when changing from near a azimuth target to one closer to the horizon?
If yes to any of the above, it is quite likely to be differential flexure somewhere. You'll probably find PHD is working fine.
I had similar issues, and eventually , after looking at the a range of complex causes, found it was simply 'flex' in my guide camera chain. When everything was stiffenned up (and balanced nicely), the problems reduced significantly.
HTH
|