Re: [CAD_CAM_EDM_DRO] Re: AMAZING STUFF!

Feedback from a DRO isn't strictly necessary to compensate for leadscrew
error, but you do have to
characterize the error vs position of the nut on the screw. Given that
info for both directions of
travel, software interpolation of error lookup table data can be used to
calculate where the cutter
position really is vs what your step commands say it should be. Not a
trivial task, but it should
be possible. You also need to know how much "slop" exists along the
motion path to account for
direction reversal. So if you can use some known position standard such
as a calibrated rule on top
of say a milling table, tediously read off position with an optical
center finder in your spindle,
and do some good programming, perhaps stepper systems don't really need
a DRO feedback to be
accurate. Its a whole lot faster to spend the money for good ball
screws, but heck, anybody could
do that!

What do you guys think? Possible, completely nuts, or someplace in
between?

Bill Martin

james owens wrote:

>
> From: "james owens" <wotisname@...>

> The problem I have with lead screw wear is quit common in that the waer is not uniform over the lenth of the screw because more use is make of the centre than the ends.
>