Re: [Emc-users] Lead screw mapping

[Sepehr K. <sk...@al...>]

If you're using preloaded nuts then mapping makes sense (your doing 
essentually what Moore invented mechanically way back when). It's much 
cheaper than linear encoder, albeit less ideal. I'm still building my 
mill but plan to try to map the screws (my screws are decent rolled ball 
screws, ~0.004" per foot rated). I'd be interested if anyone else has 
had success mapping in EMC as well.

Additional question on mapping (a bit more esoteric since my machine 
won't have the bandwidth) but does EMC have the ability compensate for 
form errors?

thanks,

=====================================
Sepehr (Sep) Kiani, Ph.D.
Email: sk...@al...
http://pergatory.mit.edu/skiani
=====================================



Gene Heskett wrote:

>On Tuesday 08 November 2005 10:03, fj...@sn... wrote:
>  
>
>>I'm still a long way from getting my machine done but I was thinking
>>about screw mapping this morning for some reason.  Is there a standard
>>technique for doing this?  Can it be semi-automated using EMC?  I was
>>thinking about mounting an extra encoder on a wheel and mounting that
>>in the machine spindle so the wheel would turn when the table moved in
>>X or Y direction. You could then compare the difference between the
>>encoder on the axis you are checking and the encoder on the wheel to
>>get a continuous mapping.  Am I way off in right field on this?
>>
>>BTW-I added myself to the "Big Map" too.
>>
>>/M
>>    
>>
>
>I'm not convinced that screw mapping would be all that usefull,
>particularly when the map result may be at odds with itself depending
>on the direction of travel, due to poor making of the screw & nut, wear
>causing backlash that varies according to where you are on the
>screw, etc.
>
>Thats the problem I faced with a micromill when finding the z axis
>screw had no backlash adjustment, and was also off as much as .010
>inches depending on its rotational position.  Backlash compensation in
>emc is not done very well, and setting anything over a thousandth or so
>when using steppers results in the steppers losing lock because the move
>is done too fast.  Rather than try and map a crappy screw, I'm in the
>process of replacing it with some 1/2 10tpi acme screws & nuts from
>Nook.
>
>IMO quality screws will be much better than trying to map a cheaply
>made one thats not big enough for the job so the wear rate is
>excessive.  In the process of using that crappy Z screw to make the
>replacement nut holder, I'd be willing to bet the about .010 backlash I
>started with is going to be closer to .015 when I'm ready to pull it out
>again & replace it with the new one I'm making.  I'm litterally washing
>half a cc of grey/black muck out of the original nut with fresh 5w30
>about twice per hour of running time.  The Nook stuff will be adjustable
>the way I made the nut holder and should stay under a thou once set &
>exercised to seat it in for the rest of my days since I'm already a
>senior citizen at 71.
>
>I thought of using ball screws, but that needed more little red
>wagonloads of money than I felt a 300 dollar micromill was worth.
>They are also typically quite fast and this one doesn't need the sort
>of rapids that would give...
>
>  
>