Hi Bob,
First, sorry for this getting stuck on the board and not getting approved. I'll spare you excuses - it's my fault.
Re your question about probe calibration -
I'd like to start by offering some insight into how Mach3 handles probe moves and the assumption it makes internally about them. Mach let you tell it the diameter of a probe's tip. It attempts to use this info to do what mach calls "tip compensation".
Essentially, mach knows internally the direction of a probe move; when the probe triggers, mach latches the trigger coordinate (in machine coords). The probe tip diam is used to adjust the trigger location to the (presumed) center of the probe tip). In doing this mach makes some assumptions: It assumes all probe tips are perfectly spherical and it assumes that the physical object that touched the probe was exactly on the center of the path being transversed when the trigger event happens.
Consider a large diameter probe - let's make it really large for sake of discussion: 2 inch diam. Now consider the center line if the probe path attempting to pass an object that is 0.5" away from the path.... the probe gets triggered as the object deflects the large probe tip - but the trigger point assume by mach is off because the object "found" is not a flat face, perpendicular to the probe path etc. This introduces error in the mach probe tip comp calculation and that error is reflected in the reported trigger coordinates. Rule of thumb: use the smallest practical probe tips.
Now i need to talk about doing probing routines in a Mach add on (like MSM). The MSM probe routines are implemented by sending mach G31 motion commands. However. all tip compensation calculation is done internal to mach. MSM basically uses the approach from wherever we are, probe to "there" using G90 (so Mach is just sent the target destination). MSM does not attempt to do complicated things such as calculate the direction of a probe move and use that to interpolate tip comp from an direction dependent error map for the probe. Frankly, that was considered too complex to be worth doing as an MSM feature. IN the early days of MSM implementation, I did play with the idea - but the potential increase in accuracy was not worth the effort.
What makes this hardly worth while are other factors: On the usual class of spindles in use for Mach which don't have indexable tooling holders - each time the probe is mounted in the spindle the error map would change. Each time the probe is used the spindle is unlikely to be at the same orientation as the last time - and hence run out errors also vary with spindle rotation position. This is why the manual suggests that one create some way to always put the probe in the spindle with a uniform spindle orientation and probe to spindle angular relationship. That minimizes some error sources.
MSM then provides a way to probe a known bore size. The resulting measurement is uses to calculate cosine errors. MSM calcs an "effective" probe tip - i.e. the probe tip which when used in the tool table, gives the correct, pre-known result for the bore size.
When all is said and done, MSM's probe results depend on mach's calculations and assumptions, probe tip size and mechanical imperfections in the machine.
I hope this info helps you decide whether MSM is an appropriate tool for your usage.
Dave
bainbob wrote:How are you supposed to calibrate your probe? After playing with this package, the probing routines are impressive, but I see no means to calibrate the probe. Do you rely on the mechanical calibrations of the probe? unless it's a Renishaw (very expensive) I've never seen a probe that was perfect. Most aren't very good at all. The other probing routines I have used create a probe calibration map and use it to compensate for errors (besides alignment there's also trigger distance (and others). As an example, I have one probe which is within +/- 0.0001 in three quadrants but off by +0.0012 in the 4th quadrant. How do you handle this? I really like your package (still in trial phase waiting for an answer). I posted this same question before thanksgiving but it still isn't on this forum (why???).
Bob Bain