- I have but I have no actually used it yet.
IMHO: There is no reason it should not work. Be advised that I have no power
crossfeed, thus, no slot. This makes it a whole lot easier but I wouldn't
hesitate to try one if required. I think the "precision" thing is blown
waaaay out of proportion. In the first place, the mechanics of the half-nut
operation tend to minify any imprecision of the screw thread and limit the
use of any "excess precision."
This is not to say that just any CPOS would be suitable but there is no real
advantage to paying extra for ultra-high precision. I tested the accuracy of
my Enco-supplied stock by ordering some of their matching threaded nuts. I
put two nuts on the rod in jamb-nut fashion and tightened them so that there
was a noticable drag when rotated as a unit. I secured the nuts to stay in
that relationship, then hand-turned them the length of the section I
expected to use. I could feel no significant variations in the torque
required to move the nuts so I judged it to be suitably accurate for my
This is a highly un-scientific testing method but I believe it to be
entirely practical, cheap and easy. The next piece of rod I check may not do
as well and, therein lies what I percieve to be the benefit of higher
precision stock.... lower reject rates in the manufacturing process. I am
perfectly willing to sort out a "good section" from a 6 feet long rod (or
two) and condemn the remainder to use in future, less critical uses. Joe's
Lathe Mfg. Co can't afford to do that nor can they afford to get a lathe
returned because the leadscrew causes a tight spot.
CAUTION: This is my 2 cents worth and it may not be worth the asking price.
>Has anyone ever made a feed screw from acme thread stock? It looks
>doable to me. A key slot on one end and the other end turned to the
>bearing diameter and threaded. Am I overlooking something real
>obvious? Thread stock is readily from McMaster Carr. Nuts are
>available in grey iron, bronze, and brass. Which would be the best
>choice for half nuts?
- At 11:12 AM 6/2/2003, WDSmith wrote:
>expected to use. I could feel no significant variations in the torqueIt's everything you say, but I'm not entirely convinced it will give the
>required to move the nuts so I judged it to be suitably accurate for my
>This is a highly un-scientific testing method but I believe it to be
>entirely practical, cheap and easy. The next piece of rod I check may not do
>as well and, therein lies what I percieve to be the benefit of higher
answers you seek. In a leadscrew positioning system there will be some
amount of cumulative error taken from one end of the range to the
other. You could graph it as a line with some slope, and on a computerized
system you could arrange to apply an opposite correction by calculation.
Then at any given point along the screw there is an instantaneous deviation
from that line ranging from zero to some maximum. If this has a cyclic
component a computer algorithm could deal with that part, but not the
rest. Not, of course, that the computer stuff is likely to happen with an
Atlas lathe -- that's for stuff that works in micro-inches.
I think your nut test might give an idea of the amount of instantaneous
error -- but I don't have a feel for how sensitive it might be, and there
would certainly be phase interactions between the length of the nuts and
the rate of change of the error.
Anyway, having said all that I haven't yet found a number for the typical
errors in a common Acme-thread drive screw, and of course your needs might
vary a lot depending on what -- and how long -- you're making, and what
class of fit. I'm guessing that for a lot of work a cumulative error of a
thou per inch and instantaneous of +/- a thou would be fine, but I don't
have enough experience to have a feel for it. Atlas's suggestion to use a
44-52 change gear combination for metric gives an error of six tenths per
inch, and they call it "very near" the correct ratio, for what that's worth.
I found a spec of three tenths per inch for some Acme "precision"
leadscrews of the sort of size we'd be interested in, and if I interpret
the Sherline FAQ page correctly their leadscrews are good to that figure
(they say "99.97% accurate). Velmer make two grades of leadscrew,
"Standard" at six tenths per inch, and "Precision" at 1.25 tenths per inch
(7 -- 1.5 thou/10"). Kerk will supply screws in the range of six tenths to
one tenth per inch. Nook work to a uniform 2.5 tenths per inch (3
thou/foot). None of them distinguish between the types of error, at least
on their websites.
I've been unable to find even a picture of a Genie drive screw, let alone a
spec for it.
David Beierl -- Providence RI USA
Atlas 618 6"/3" lathe ca. 1941
- Pictures I can get. Just let me get the new opener hung and working. I'm
having to repair someone else's poor installation of the mounting hardware.
----- Original Message -----
> I've been unable to find even a picture of a Genie drive screw, let alone
> spec for it.