Re: Climb Milling
Posted by
ballendo@y...
on 2001-06-05 20:26:41 UTC
Buck,
Your message reminds me of yet another "gotcha" in IMPLEMENTING
software backlash comp.
What do ya do when the next axis reversal is a move LESS than the
backlash amount? Or a series of moves WITHIN the backlash amount of
the machine. Or even worse, a series of short moves in/out and around
the "backlash band". I've seen software backlash compensation
algorithms which would take your mind quiz and FAIL it!
BTW, with a sherline advocated "normal" backlash of .003, and CAM
generated toolpaths(often lots of short line segments), this is not
a "theoretical" question...
To answer the mind quiz: NO, backlash does not "add up". The
exception is what I mentioned above; i.e., a control which does not
implement compensation correctly CAN make the b/l "seem" to increase.
Hope this helps.
Ballendo
P.S. It would be a good idea to repeat your backlash measurement
routine at several locations along each axis; and also using
differing lengths besides one inch. And lastly, a drilled hole will
not be a good thing to measure to, on any but the most basic (and
likely low-resolution) machines.
Your message reminds me of yet another "gotcha" in IMPLEMENTING
software backlash comp.
What do ya do when the next axis reversal is a move LESS than the
backlash amount? Or a series of moves WITHIN the backlash amount of
the machine. Or even worse, a series of short moves in/out and around
the "backlash band". I've seen software backlash compensation
algorithms which would take your mind quiz and FAIL it!
BTW, with a sherline advocated "normal" backlash of .003, and CAM
generated toolpaths(often lots of short line segments), this is not
a "theoretical" question...
To answer the mind quiz: NO, backlash does not "add up". The
exception is what I mentioned above; i.e., a control which does not
implement compensation correctly CAN make the b/l "seem" to increase.
Hope this helps.
Ballendo
P.S. It would be a good idea to repeat your backlash measurement
routine at several locations along each axis; and also using
differing lengths besides one inch. And lastly, a drilled hole will
not be a good thing to measure to, on any but the most basic (and
likely low-resolution) machines.
--- In CAD_CAM_EDM_DRO@y..., yahoo@a... wrote:
> To Measure backlash:
> Travel one inch in the positive direction;
> Drill hole number one.
> Travel one more inch in the positive direction;
> Drill hole number two.
> Travel back 1/2 inch, in the negative direction;
> Drill hole number three.
> Hole three will not be 1/2 inch back from hole number two, by the
> backlash amount.
>
> Yes, it is used everytime a motor reverses direction. It applies to
> circles - everything.
> The motor runs an extra amount, equal to the backlash, for only the
> first move after each direction change.
>
> The extra motor movement is just "taking up the backlash".
>
> Backlash can be a combination of gaps and defections (under load).
> Gaps are symetrical, deflections may not be and defections are
> usually somewhat (nonlinear) force dependent.
>
> A mind quiz: If you have a back and forth moves, does the backlash
> accumulate? Or after 10 back and forths are you off from where you
> started by 10 times the backlash?
>
> Buck
Discussion Thread
ron ginger
2001-06-04 15:35:10 UTC
Re: Climb Milling
ballendo@y...
2001-06-04 17:57:33 UTC
Re: Climb Milling
dougrasmussen@c...
2001-06-04 20:28:45 UTC
Re: Climb Milling
Jon Elson
2001-06-04 21:59:59 UTC
Re: [CAD_CAM_EDM_DRO] Re: Climb Milling
yahoo@a...
2001-06-05 05:32:00 UTC
Re: Climb Milling
Alan Marconett KM6VV
2001-06-05 10:58:00 UTC
Re: [CAD_CAM_EDM_DRO] Re: Climb Milling
Jon Elson
2001-06-05 11:30:52 UTC
Re: [CAD_CAM_EDM_DRO] Re: Climb Milling
Hugh Currin
2001-06-05 12:30:11 UTC
Re: Climb Milling
Rich D.
2001-06-05 12:38:26 UTC
Re: [CAD_CAM_EDM_DRO] Re: Climb Milling
Fred Smith
2001-06-05 15:14:44 UTC
Re: Climb Milling
RC
2001-06-05 15:35:17 UTC
Re: [CAD_CAM_EDM_DRO] Re: Climb Milling
dave engvall
2001-06-05 16:13:45 UTC
Re: [CAD_CAM_EDM_DRO] Re: Climb Milling
dave engvall
2001-06-05 16:16:45 UTC
Re: [CAD_CAM_EDM_DRO] Re: Climb Milling
ballendo@y...
2001-06-05 18:57:23 UTC
Re: Climb Milling (again)
Smoke
2001-06-05 19:22:34 UTC
Re: [CAD_CAM_EDM_DRO] Re: Climb Milling (again)
Sven Peter, TAD S.A.
2001-06-05 19:29:40 UTC
Re: [CAD_CAM_EDM_DRO] Re: Climb Milling
yahoo@a...
2001-06-05 19:46:12 UTC
Re: Climb Milling
Tim Goldstein
2001-06-05 19:50:35 UTC
RE: [CAD_CAM_EDM_DRO] Re: Climb Milling (again)
dougrasmussen@c...
2001-06-05 19:53:22 UTC
Re: Climb Milling (again)
ballendo@y...
2001-06-05 20:10:17 UTC
Re: Climb Milling (again)
ballendo@y...
2001-06-05 20:26:41 UTC
Re: Climb Milling
ballendo@y...
2001-06-05 20:29:03 UTC
Re: Climb Milling (again)
ballendo@y...
2001-06-05 20:33:01 UTC
re:backlash comp was Re: Climb Milling (again)
Chris Stratton
2001-06-06 04:48:36 UTC
Re: [CAD_CAM_EDM_DRO] re:backlash comp was Re: Climb Milling (again)
ballendo@y...
2001-06-06 05:25:57 UTC
Re: Climb Milling
yahoo@a...
2001-06-06 05:32:56 UTC
re:backlash comp was Re: Climb Milling (again)
ballendo@y...
2001-06-06 06:18:16 UTC
re:backlash comp was Re: Climb Milling (again)
Smoke
2001-06-06 12:05:16 UTC
Re: [CAD_CAM_EDM_DRO] Re: Climb Milling (again)
Smoke
2001-06-06 12:05:38 UTC
Re: [CAD_CAM_EDM_DRO] re:backlash comp was Re: Climb Milling (again)