"Jim Wilkins" wrote in message
...
On Nov 28, 1:14 pm, "Ed Huntress" wrote:
"Jim Wilkins" wrote in message

...You could file the left-hand
bearing seat on the spindle to a moderate press fit so it can slide if
it has to,...
Right. But if the bearing slips, you have no Z-axis stiffness. The spindle
would slide in or out when you took facing cuts or faced the back of a
bearing retainer or whatever on a shaft.

The traditional setup, with two facing, angular-contact bearings at the
front, and one floating bearing at the rear, solves all of those problems.
There isn't enough space between the front bearings for thermal growth to
be
a problem. And both thrust loads, Z+ and Z-, are taken out at the front.
Then the rear bearing can float a bit with no loss of axial stiffness. It
needs to be preloaded in the radial direction to maintain radial
stiffness,
but there, too, there is more room for slack than at the front.

jsw

There's no conflict, I suggested a non-precision manual way to make
the rear bearing float without being too loose when you are boot-
strapping the lathe. I'd figure out the preloading details on the
chuck end after a trial assembly to see if the bearings are good
enough. Some combination of easily made spacers and shims ought to
work.

How do you take out both Z+ and Z- axial loads from one single-row bearing?

--
Ed Huntress


I've found that if I order the parts to build it one way, perhaps not
the absolute best one, they will allow it to be assembled several
other better ways that I think of only after trying the first one. I
do have to know any hard limits like bearing PV and buy to meet them,
but the mounting details can safely wait.

On our plain bearing, leather belt driven South Bends a ball thrust
bearing on the pulley side of the rear/left spindle bearing absorbs
thrust and a threaded shaft clamp on the outside takes up play. The
instructions are to hand-tighten it, then back off 3/8" (of rotation)
and lock the clamp screw. The chuck end is free to slide.

jsw