Two different values when ran in AutoRun and normal Calypso.

[[Ri...]]

This doesn't make any sense at all.

One program, it is setup in AutoRun.

It has some PCM to determine if the part is tilted in the X axis.

measure("YZ Finder")
TRYAGAIN:
measure("Tilt Finder A")
measure("Tilt Finder B")
if abs(getActual("Tilt Finder A").x-getActual("Tilt Finder B").x)>0.004
message("Please Re-Position part.  Ensure that the Flange of the part is fully to the left of the V Block and X Bore is straight down")
goto TRYAGAIN
endif

When ran in Autorun it reports .00432xxxxx. (Gets stuck in an endless loop due to false value).

When run in Calypso it reports .00234xxxxx. (Works as written).

Part is mounted in the fixture and secured, and I can get the same results at will.

Same program, same points, the program doesn't change, only the result of the actual probed points.

Tilt Finder A and Tilt Finder B.

Part is cylindrical and its axis is in X. Plane surface is ZX.

Tilt Finder A touches a plane surface at Y0, X0 and Z+.945

Tilt Finder B touches a plane surface at Y0, X0 and Z-.945

How is this possible?

 

[[Da...]]

Exactly the same basesystem selected in the CNC start settings of Calypso and AutoRun? Are you using a pallet alignment system?

Screenshots of both CNC start settings would be helpful.

[[Ri...]]

Base System doesn't change.

The pallet is a single part.

CNC start settings are identical.

Edited September 5, 2025

[[Cl...]]

Have you tried deleting the program from autorun and then re-loading it?

[[Ri...]]

Please sign in to view this quote.

No, I have not, not sure what that will do but I will try it.

This issue is on both CMM's that utilize Autorun for families of parts, I suspect it has been an issue for many years, and no one noticed.

[[No...]]

Out of curiosity: Are there any patterns involved in your program?

[[Ri...]]

Yes, too many to count.

The patterns are all circular and in XZ.

The points A and B are only in the X axis.

[[No...]]

And are the deviating results connected to specific elements of a pattern or is it all random?

In my case it was always the first element of a 10-fold linear pattern. All the others were identical.

[[Ri...]]

I don't think I understood your question initially.

The two points are not patterns, but I do use a lot of patterns for other features.

The points touch a plane, one at +Z and the other at -Z, this is to detect if the part is tilted in the ZX axis.

The points are after the Base Alignment.