Dave,
I'm going to guess that they didn't put external encoders on the ends of the screw on this model. If there is an external encoder in addition to the one inside the red cap motor, disregard the following. We'll have to wait until I can look it up for you Monday.
Generally, the idea in zeroing an 18T, is to rapid up to the decel cam, slow down, fall off the decel cam, watch for the next marker pulse and stop. Some things have to happen first however: the axis must be up to zeroing speed (defined as 128 counts of servo lag) before it sees the marker the first time, and it must have seen a marker before it sees the decel switch (remember it sees one every time the motor turns). If the motor is oriented wrong, it might overtravel before it finds the first marker after falling off the switch.
On a Hardinge, the brute force fix for this might just be to loosen the motor, turn the motor shaft a little and try again. Not too scientific, but effective.
After a true zero return, the control looks at the grid sync parameter (if it is non zero) and jumps to an exact point. This is useful on a Hardinge to get the exact distance from centerline that they advise.
Good luck,
Warren
www.uptimecorp.com