I had the same trouble a few months ago and decided to machine a block of
steel to a known accurate surface, then using digital readouts use a
centre-finder to see if it would correctly indicate the centre. It varied by
about 0.2mm ( about 8 thou ) through about 10 trials. First I used the
methode of slowly creeping to wards the edge until the end went still (
concentric ) and then tried going past this point until it suddenly flicked
out. Both methods didn't seem to offer any better reliability. So I went
back to my old ground steel bar held in an autolock chuck. That never lies.

Dean.

"David Anderson" wrote in message
igy.com...
What is the proper technique for using an edgefinder in a vertical milling
machine? For this example, I'm using one with a 0.200 inch diameter
projection, thus 0.100 inch offset from true centerline. Prior to making
contact, the 0.200 extension is off center, but a soon as contact with the
work piece is made, it seems to be now rotating true (machine is running
under power about 150 rpm). Ok, now set x or y graduated dial to zero.
BUT,
what happens if I advance the table, say another couple of thousands? How
can I be sure that the first contact is correct if I can still move the
work
a tiny bit closer to the indicator without anything happening? Tool made
in
India (no instructions); Machinery's Handbook says nothing about
edgefinders. HELP!