Aha! As I think bernard said when he first mentioned it, it takes only the tiniest gap in the magnetic path to have an enormous effect - the two parts of the magnetic core must be, and remain, in contact.That was it John, even a fraction of a mm causes a massive drop in the reading! The extra bit of sheath, rope, etc... in the sensor must of pushed open the clamp, only by 10th or so of a mm, but that's all it took.In that case, is not by far the most likely explanation the one that was given right at the start - namely that putting additional things through the sensor is stopping in 'closing' (and hence completing the magnetic circuit) properly?
Kind Regards, John