Standard deviation in probe qualification

[[Ra...]]

What are potential reasons for the fact that I'm getting a standard deviation a way out of specification (.0011) while doing automatic probe qualification?
It's a Vast XT probe sensor and I'm using a angled calibration sphere..

[[Ma...]]

- Dirt or debris
- Damage to stylus or calibration sphere
- Loose or damaged stylus
- Damaged mounting plate
- Probe head
- Air bearing(s)
- Self-Leveling system
- Drive bands (some machines)
- External vibration
- Anomalies in electrical or air supplies\
Probably some others I missed

[[No...]]

Please sign in to view this quote.

...like abrasion from scanning a lot of rough material

- Long and thin stylus shank -> tends to larger deviation

[[Ow...]]

If it is 0.0011mm, it's not way out of calibration.
If it is 0.0011 inches, then yes, something is not right.
Along with previously mentioned:
*Carbon-fiber shaft probes can get a crack in them that the naked eye will never see.
*Loose ruby or calibration sphere can come loose.

[[Aa...]]

We mostly scan parts with square, flat faces, so whenever our standard deviation starts creeping up, we know we're developing "flats" on the ruby. If we examine it under the microscope and shine a light on it from just the right angle, we can see glare off the flat face. The thing to bear in mind is that the way standard deviation is calculated has a tendency to mask how bad the flats are. (Our old microscope's toplight was exactly the right angle to spot the flats.) The thing to bear in mind is that the way standard deviation is calculated has a tendency to mask how bad the flats are. We used to throw the ruby styli away and replace them, but we determined we could save money by using a diamond stylus.

[[Aa...]]

Also, it's worth noting that your Reference Sphere could also be developing flats. You could test this by unthreading it from the base and putting a washer or shim ring in between so that the RS is oriented differently. I think part of the reason we develop flats here is that our parts may have diamond polishing compound residue on them when we scan them. But if anyone is scanning parts made of especially abrasive material, that in itself could be the culprit.

[[Er...]]

Our calibration sphere is one that has no adjustment. Turned out the epoxy holding the sphere had degraded.

After a few hours of trying to qualify different probes I decided to "aggressively clean" the sphere. It was then I noticed something wasn't right. The sphere wasn't what I would call loose. It just had an infinitesimal amount of play. We were able to remove the sphere and re-glue it.

[[Th...]]

What is the acceptable range for Sigma deviations?

[[No...]]

Please sign in to view this quote.

There is no fixed value, as it depends on how long and thin the stylus is.

[[Is...]]

We set the Std Dev. limit in 0.0005mm (0.0003mm for new probes)... most of our manufacturing tolerances are between 0.010 - 0.030 mm

stddev.png