Re: [CAD_CAM_EDM_DRO] Making leadscrews more accurate
Posted by
Les Watts
on 2001-01-05 13:29:53 UTC
> Les, I understand from your write up (very well done!) how your making theI
> lead on the second ball scre match the lead on the first ball screw. What
> don't see is how your making the pitch of either lead screw more accuratefirst
> than it was originally. To do that I would presume you would have to use
> the same cam tecnique on BOTH screws. You'd have to adjust the cam on the
> first screw to "iron out" all the irregularrities (pitch errors)on the
> screw and then adjust the cam on the second screw to match the travel ofthe
> first screw. Correct?That's exactly correct.
>
> Smoke
Note that when the pair is matched I use electronic means
to map out the absolute error. That's just a look up table
(really a power series) that sends the axis to a slightly different
coordinate to really go to the right place. It's preprocessing
so it uses up no servo resources or phase margin. It can be done separately
from the control program.
But it will work just as well with two cams (or one on a single
leadscrew) to correct absolute errors just as you say.
The key to doing it cheaply (without an interferometer) is by
stacking inexpensive micrometer calibration bars like jo blocks
and measuring the gap left with an indicator.
The cost of high accuracy class ground screws goes up
really disproportionally with the accuracy you get. I can get
the same results at a tiny fraction of the price with rolled
antibacklash ballscrews and this technique.
I figured this has been used some since the early industrial revolution but
I had not seen it elsewhere. And I think it was
Maudslay that used it around 1820 or so. As I recall his lathes
had dials calibrated in tenths in a time when there was no prime machine to
generate an accurate screw lead directly.
There are other ways (repetitave averaging) if one is creating
a pitch standard from scratch but the cam method converges MUCH faster.
Leslie Watts
L M Watts Furniture
Tiger, Georgia USA
http://www.rabun.net/~leswatts/wattsfurniturewp.html
Discussion Thread
Les Watts
2001-01-05 11:00:26 UTC
[CAD_CAM_EDM_DRO] Making leadscrews more accurate
Drew Rogge
2001-01-05 11:43:08 UTC
Re: [CAD_CAM_EDM_DRO] Making leadscrews more accurate
Smoke
2001-01-05 11:52:06 UTC
Re: [CAD_CAM_EDM_DRO] Making leadscrews more accurate
Carlos Guillermo
2001-01-05 12:17:54 UTC
RE: [CAD_CAM_EDM_DRO] Making leadscrews more accurate
Les Watts
2001-01-05 13:29:53 UTC
Re: [CAD_CAM_EDM_DRO] Making leadscrews more accurate
Smoke
2001-01-05 15:22:34 UTC
Re: [CAD_CAM_EDM_DRO] Making leadscrews more accurate
Ian Wright
2001-01-05 15:37:14 UTC
Re: [CAD_CAM_EDM_DRO] Making leadscrews more accurate
Les Watts
2001-01-05 16:10:43 UTC
Re: [CAD_CAM_EDM_DRO] Making leadscrews more accurate
Jon Elson
2001-01-05 16:19:20 UTC
Re: [CAD_CAM_EDM_DRO] Making leadscrews more accurate
Smoke
2001-01-05 21:24:16 UTC
Re: [CAD_CAM_EDM_DRO] Making leadscrews more accurate
ballendo@y...
2001-01-06 15:45:47 UTC
re:Making leadscrews more accurate
ballendo@y...
2001-01-06 15:50:13 UTC
Re: Making leadscrews more accurate