Why didn't this work?

There was a thread recently about cheap drill chucks. I commented that I had a cheap Jacobs with bad runout. That got me to thinking that I could fix the runout. Here's what I did:

- chucked up a piece of steel and turned it to 7/16+- (an arbor)

- fastened the drill chuck on this arbor by tightening its jaws on it

- faced the (mounting) end of the chuck and bored it .600+-, about 3/8 deep

- chucked up another piece of steel & (right to left): - turned 1/2-13 threads for about 1/2" (the drill chuck mounting threads) - turned to .600 for almost 3/8" (a register diameter) - faced a shoulder (register face)

- mounted the drill chuck on this new arbor

- chucked a short piece of drill rod in the drill chuck

The runout on the piece of drill rod was .025 - .030 ('don't remember exactly)!! WHAT?? How could that be? Why wouldn't it be dead on?

This is how some lathe chucks are mounted (with a register diameter and face).

What am I missing?

Thanks, Bob

Thanks for all the replies. A couple of points were suggested, but were covered. The drill chuck was new, so it wasn't worn. The 2nd arbor was not removed from the lathe before checking the runout. The size of the drill rod was close, but not exactly the size of the 1st arbor.

I've been thinking about this and it seems that the procedure should be much more accurate than I'm getting and that the problem is likely in the skill of the operator. Particularly in turning the register diameter of the 2nd arbor. Any undersize there could show itself in the TIR. Although it seems quite unlikely that it would be .025 under!

I going to redo the 2nd arbor keeping in mind the points raised here and being as careful as I can with the register diameter. I'll post a follow-up.

Thanks again, Bob