What's new
What's new

Renishaw probe; best way of finding center of a square tilted at 45 degrees?

Can you M19 R180 the spindle so the probe is hitting from the same quadrant, just turned around? Would that comp out the error?
There is a setting in the master probing cycle to automatically reorient the probe so it will spin to deflect in the same orientation when running vector cycles.
(Parameter #131, Bit 10, set to 1024 in O8710 in Brother/Blum world)

Theoretically its an issue but IME its not necessary.
We probe every part we mill to find center from two sides (~100 parts a day) so it was worth the effort to do some testing and save a few seconds/cycle.

I switched from 4-point cycles, to 3-point with orientation, to 3-point without and haven't seen any real downsides. Once I switched from a 3-point vector boss cycle, to qty. 3 discrete single-point vector cycles and did the centering math myself is when I saw the most significant reduction in cycle time. Retuning to center between each touch is for suckers...

Back to back testing between 3- and 4-point cycles on the same part I would see changes of a tenth' or two but that would also happen running the same 4-point cycle several times in a row.

This is probing on finish machined surfaces, with diameters of 6-10". If probing smaller/larger and rough surfaces YMMV.
 
Last edited:
Trust exercise: Next time you're probing a hole for location, probe it twice; Once starting dead center and once starting way offset inside the hole. I bet your locations are within a couple tenths of each other.
 
Trust exercise: Next time you're probing a hole for location, probe it twice; Once starting dead center and once starting way offset inside the hole. I bet your locations are within a couple tenths of each other.
On the Mitsubishi EDM's we had at Mitsubishi, (hmmm), we changed ours to probe in triplet., 3 touches . instead of just one. Our location improved dramatically.
 
"Once starting dead center and once starting way offset inside the hole"

If you are running vector cycles and starting from the middle of nowhere sure there will be more error. Center would be calculated from some valves within +/-15 degrees of the actual displacement. But if you already know about where to start from I've never noticed enough of a difference to matter.

For the OPs application its mounted on some machined blocks, they have a pretty good idea where center is but just want to accommodate for gripping on a rough surface and make sure the ends line up. Seems to me like there's probably more "error" in the raw material surface than would be introduced using a vector cycle vs a traditional 4-point, or probing in triplicate.
 
"Once starting dead center and once starting way offset inside the hole"

If you are running vector cycles and starting from the middle of nowhere sure there will be more error. Center would be calculated from some valves within +/-15 degrees of the actual displacement. But if you already know about where to start from I've never noticed enough of a difference to matter.

For the OPs application its mounted on some machined blocks, they have a pretty good idea where center is but just want to accommodate for gripping on a rough surface and make sure the ends line up. Seems to me like there's probably more "error" in the raw material surface than would be introduced using a vector cycle vs a traditional 4-point, or probing in triplicate.
You missed my point. Starting offset in a bore is the same thing as probing two equally angled surfaces, like the OP is doing. See picture. The center (zero) found for probing A and B will be the same, within a couple tenths at most in my experience. I've tried this specifically to see how far off an automated setup can be and still get good results. Obviously this has limits, but 45º walls are nowhere close to the limits.
20240109_105736.jpg
 
What are you doing to the end of the bar? I have a live tool lathe with the capacity to hold this, so the lathe guy in me really wants to do this over there:D
 








 
Back
Top