[CAD_CAM_EDM_DRO] Making leadscrews more accurate

> Les wrote:
> >Rather than XXYZ control the minor pitch variation between the two x
> >screws is taken out by a simple cam/rotating ballnut.
> >It's an old victorian trick to make a leadscrew's pitch more
> >accurate than it was manufactured to.
>
> Les,
>
> Please tell us more about this! And/Or post links/Refs??
>

Well I have no links because I know of no other one that has done this. I do
seem to recall reading about the early machine makers in the 1800's using
this technique to make leadscrews though. (Maudsley?)

I will be happy to do a little writeup here as I will be grilling
you guys about an EMC install pretty soon! Let the mechanics
be as open source as the software.

A ballscrew or other leadscrew has pitch errors that can accumulate to large
values over a distance of feet. The preloaded rolled screws I like to use
are C7 which has
an e300 of .04mm. In other words .002"/12" pitch error.
It sometimes is periodic and cancels over length but often does not. There
is much confusion about cumulative vs non cumulative errors in ballscrew
specs. I do not think thomson
even uses the e300 spec for example. Ask a sales rep whether pitch errors
accumulate over distance or are periodic
and he will say "yeah".

Anyway when using tandem screws this error could be double that.... enough
to seriously bend a structure. Often the screws are driven by separate servo
channels individually mapped out.
But this costs $$$ and if the electronic gearing glitches...bad
bad things happen.

So I chose a simple robust mechanical approach.
One of the nuts is in a thrust bearing that allows it to rotate about 45
degrees as it moves. A ball bearing cam follower
is on an arm attached to the ball nut as well.

This arm engages a strip of steel mounted parallel to the screw travel that
acts as a cam. It is held by a spring strong enough that it will always stay
engaged.

So as the nut moves along it can rotate slightly adding and subtacting small
amounts to the travel. The rotation depends
on the shape of the cam strip.

The system is easy to calibrate because it only requires
comparative measurements. Use a dial indicator or caliper
just clamped to a long stick. Measure the difference in travel
with equal turns of both ballscrews. Record the data and then
simply bandsaw the cam strip to correct it. On mine the rough
sawn edge is good enough for the cam as 3/16 "=.001"
carriage travel.

It is very likely that two screws will have consistent but different pitch.
then the cam strip is left straight but slightly tilted.

Of course the two screws are driven by the same motor via
gears or other means.

Then the two screws can be error mapped as a pair if desired. A cheap way to
do this is just to buy a $30 12" micrometer test bar.
Replicate a few of these and stack them with a 6" or more
dial caliper on one end. The test bars are very accurate and this can be
used to map out the whole axis (at a temperature).

This saves a servo axis and preserves phase margin of the system compared to
XXYZ. It is much simpler than some kind
of articulated gantry allowed to parallelogram but still be very stiff.

Make sense? descriptions are harder without pictures!

Les

Leslie Watts
L M Watts Furniture
Tiger, Georgia USA
http://www.rabun.net/~leswatts/wattsfurniturewp.html

> >Having made the plunge from hobby to profession I am tending
> >to make too much myself and not outsource enough!!
>
> You're probably in good company here, then! :-)
>
> Ballendo
>
>
> Welcome to CAD_CAM_EDM_DRO@...,an unmoderated list for the

discussion of shop built systems, for CAD, CAM, EDM, and DRO.

>
> Addresses:
> Post message: CAD_CAM_EDM_DRO@egroups.com
> Subscribe: CAD_CAM_EDM_DRO-subscribe@egroups.com
> Unsubscribe: CAD_CAM_EDM_DRO-unsubscribe@egroups.com
> List owner: CAD_CAM_EDM_DRO-owner@egroups.com, wanliker@...
> Moderator: jmelson@... [Moderator]
> URL to this page: http://www.egroups.com/group/CAD_CAM_EDM_DRO
> FAQ: http://www.ktmarketing.com/faq.html
> bill,
> List Manager
>
>
>