This is a SPHERE 0..5/16" drill chuck. My Dad and I bought it from Paul's Tools in Sydney, in the 1970s. It never ran true. I think it came from a bargain bin, and we might have taken it back when he discovered the wobble. (they probably refunded, and said "keep it - we can't sell it")

He kept it for 40 years, and I couldn't bear to throw it away.
When I restored the Hercus, I bought an MT2->3/8" arbor to use it in the tailstock.
It is sort of OK, but not for accurate drilling.

Today is the day I investigate the wobble!


1) Throw it in an MT3 sleeve in the headstock, and measure TRO:
IMG_3801.jpg IMG_3802.jpg IMG_3803.jpg
so, 40mm from the arbor, .25mm. 50mm = .27mm. 100mm = .46mm.
It seems to be a parallelism issue?


2) The arbor thread is a loose fit in the chuck, until the last little turn (maybe 20°).
The shoulder on the arbor squares up on a little flat around the thread of the chuck.
I wonder if I can surface grind it?
Put the little front of the chuck on the mag plate, with the high point facing me,
and carefully skim 10 or 15 microns off:
IMG_3804.jpg IMG_3805.jpg
Note that the back is little low, which is the opposite of expected,
so probably not the cause of the non-parallelism?


3) Re-measure:
IMG_3806.jpg
.16mm? That is a little better than .27, but not good enough.
Interestingly, the high spot has rotated around on the chuck a little.


4) Time to hack a tiny angle. Very thin paper (a sheet from a restaurant order book) under the back:
IMG_3807.jpg
...and the test...
IMG_3808.jpg

Yay!!! 40 microns? That is an 80% improvement!!!


5) The real test is the jaws gripping something:
IMG_3809.jpg


Not quite as good. 0.14mm

which is better than initial TRO, but the end of the rod - which is 300mm long - is visibly wobbling.


I could use thicker paper and grind it again, but I'm tempted to leave it at that??
Attached Images




Read the full thread at metalworkforums.com...