Millrite X-axis thrust bearings were frozen
Joseph Gwinn
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
Too_Many_Tools
Good analysis.
Great repair.
TMT
John B. Slocomb
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.
John B. Slocomb
Joseph Gwinn
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
John B. Slocomb
<joeg...@comcast.net> wrote:
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)
Joseph Gwinn
John B. Slocomb <johnbs...@invalid.com> wrote:
> On Mon, 29 Mar 2010 00:02:04 -0400, Joseph Gwinn
> <joeg...@comcast.net> wrote:
>
> >In article <tk80r5lgc9703i0qm...@4ax.com>,
> > John B. Slocomb <johnbs...@invalid.com> wrote:
> >
> >> On Sun, 28 Mar 2010 20:12:33 -0400, Joseph Gwinn
> >> <joeg...@comcast.net> 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
> >> > ....
[big snip]
> >>
> >>
> >> 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.
>
> 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.
Yes, there was play. Two things were wrong.
First, the saddle-knee gib needed to be tightened. This solved the
cut-in-one-direction problem. As it happened, it was climb that didn't make
contact, but climb-versus-conventional was not the issue - the problem was due
to the table cocking one way then the other way. Self-feeding was not the
problem, for lack of contact.
Then, the thrust bearings at the left end of the table were cleaned, greased,
installed, and preloaded. This greatly reduced the difficulty in setting the
table to a specified X-location on the DRO.
Next will be to tighten the table-saddle gib, but this is a big deal (as removal
of the table-saddle assembly is required) and so it may be a while.
Joe Gwinn