I do this routinely every time I am out testing scopes and mounts. A star's position will change in the sky from beginning to end. The RA/Dec will change, so that the reported coordinates after a plate solve will come out different. I see this every night that I am imaging. And it has zero to do with any cable snags or other mechanical issues. It is pure and simple atmospheric refraction that is bending the tracking path of an object.
Roland, I sent you a private email, but when auto-guiding, the guide star is usually not far from the area being imaged, so the guide star is *also* refracted. Thus the only refraction effect is the difference between the refraction seen at the guide scope and the main scope, which is usually pretty small.
