Re: [CAD_CAM_EDM_DRO] Sherline CNC retrofit

In a message dated 12/9/2002 12:09:58 AM Central Standard Time,
tictocdoc@... writes:

> I noticed that the Sherline CNC upgrade includes "ball screws". I am not
> familiar with this term. Can anyone explain to me why I should upgrade to
> ball screws?

Michael:

You know what BALL-bearings are. "Ball Screws" simply means that hardened
BALLS "connect" or "transfer" the energy from the nut to the screw. Instead
of "V" or "ACME" threads, the screw has semicircular "threads", actually
usually "gothic arch" shape (comlicated story!), almost half as deep as the
dia. of the balls. The "pitch" must be a little more than the max. dia. of
the balls used. Most "machine lead-screws" have "preloaded" ball-nuts,
meaning that the "nut" has TWO "sets" of balls in separate "nut housings"
which are fastened together with a usually-adjustable "preload device" which
can "spread these halves apart" so that the balls of ONE set press on the
RIGHT side of the threads in the screw, and the balls of the OTHER set press
on the LEFT side of the threads in the screw. This is called "pre-load", and
results in TOTAL loss of "backlash" or "play". (But it DOES make the "nut
assy." long and often cumbersome!) That is, a fraction of a degree of
rotation causes an unmistakeable linear movement; it is not necessary to turn
the screw several degrees until the linear movement "reverses", as is
necessary on even the finest ACME-screw setups (lathes, mills, etc.).

Retrofitting an otherwise-good mill or lathe with good, preloaded ball-screws
simply means any future rotation of the screws will be PRECISELY converted
into the intended LINEAR motion. For smaller-scale machining operations such
as one would find in clockmaking, preloaded ball-screws, seems to me, would
be absolutely necessary, unless the operator had enough years of experience
(and time to DO the work!) to always "begin and do a cut" in the same
direction, every time. This is not always practical, and not even the "best
idea" for CNC work, anyway!

What have I forgotten? Upon re-reading your questions: For the "column"
(Z-axis on a mill), perhaps an ACME screw alone, without a preloaded nut,
would be OK, as GRAVITY provides the pre-load. However, the steppers or
servo-motors and drives you choose for the Z screw will have to be sufficient
in power to handle the weight of the quill-head AND the friction of the ACME
screw, whereas a good ball-screw can have efficiencies approaching 95% or
even more.

For the rotary table, I imagine that would have spiral- and/or worm-gears in
it, rather than lead-screws. I used one, once, but I do not have one in my
own shop, so haven't a "feel" for what that would involve. But a "GOOD one"
might not need the preload you would want in the mill's table-screws (X- and
Y), I would guess.

At least in MY "Bridgeport Knockoff" (that's a Taiwanese
machine-manufacturer!) there are "backlash adjustments" on the
leadscrew-nut-yoke, and "the book" suggests one adjust the backlash to about
0.005". No, it did NOT come "factory-adjusted" any better than about 0.02" !
! !). In fiddling with that, I discovered that any attempt to adjust the
backlash to, say, only 0.002 or 0.003" resulted in NOTICEABLE variable
friction, depending upon "where on the lead-screw" the table was at any
moment! I.e., NOT atomic-energy-precision from the source!

I hesitate to use THIS message as a place to describe my own experiences with
good ACME screws with preloaded Turcite nuts, both from BSA in San Jose; but
if more than one or two of you wanna READ more on this, actually kinda O.T.,
lemme know! Actually, come to think on it, as this "backlash and
ball-screws" has SPECIFICALLY to do with CNC technique, I say it really ISN'T
O.T. now, is it! Wot?

Jan Rowland


[Non-text portions of this message have been removed]