On Mon, 29 Mar 2010 00:02:04 -0400, Joseph Gwinn
wrote:

In article ,
John B. Slocomb wrote:

On Sun, 28 Mar 2010 20:12:33 -0400, Joseph Gwinn
wrote:

On my old (built in 1965) Millrite MVI vertical mill, one annoyance is the
difficulty setting the table to a specified X (side-to-side) location. One
would overshoot, in both directions. My first theory was that this was
because
the X-axis feeedscrew and nut were worn, but I wondered if there was an
adjustment needing attention, so I got a bright light and looked at the
screw as
I cranked the table back and forth. I could see that things were moving a
bit
at the left end, where the leadscrew is fixed to the table by two opposing
thrust ball bearings, so I decided to take this assembly apart and de-spooge
it.

This assembly has never before been disassembled, I think. The three hex
socket
cap screws that hold the left end "bell" to the table were glued in place by
well-hardened spooge, and took some force to disassemble, and there were
lumps
of spooge everywhere. Oddly, the deep blind threaded holes into which these
cap
screws go were filled with oily swarf. Cleaned everything out.

It turned out that one of the thrust ball bearings was caked with hardened
grease, being unable to move, and so the whole assembly rotated, forcing the
preload nut to turn and loosen, so everything could move. This was a major
cause of lost motion.

Soaking the two thrust bearings in acetone for four hours dissolved the
caked
grease and allowed free rotation and full cleanout. Bearings looked OK, so
packed them with Mobil-1 synthetic grease and reassembled the end assembly.
The
preload nut is round and engages threads machined into the feedscrew,
rotation
being prevented by a setscrew in the nut. I didn't like the idea of mashing
a
cup-point setscrew down on and mangling those threads, so I put a little
piece
of annealed copper in under the setscrew, so the copper would mash and
conform
to the steel threads. I tightened the nut first by hand, and then by an
added
1/8 turn with a pin spanner wrench, only then driving the setscrew home, so
there is a slight drag when cranking the screw.

The lost motion is now about 0.025", which isn't much better than what it
was
before, but it does seem easier to set the table to a desired location.
I'll
use the mill for a while and see if this is really true. With a DRO, the
real
issue is ability to set to a desired location, not the accuracy of the
screw.
Both the leadscrew and matching nut (5 tpi) are made of steel, but the nut
(being far shorter) should have suffered more wear than the screw.


Joe Gwinn


The usual procedure is to make all final adjustments from one
direction which negates any looseness in fit between the lead screw
and the nut. Most old guys do it without thinking - back it off a half
turn too far and go back 'till you get where you want to be for the
next cut.

Yep. Though I'm still trying to achieve true old-guy machinist status. I had
come to the go-one-way solution, but cutting in one direction but not the other
(even with the X-axis clamped) still caught my attention.

Joe Gwinn

I apparently didn't read about cutting in one direction but if that
happens there must be some play somewhere. Climb milling will
certainly pull the work into the cutter but that is usually obvious
when it happens and of course, if there is play somewhere, the
opposite cutter rotation will push the work out of the way.

John B. Slocomb
(johnbslocombatgmaildotcom)