Bizarre readings in PC-DMIS

You need to set your alignment correctly.  If your using 3d to measure then whatever your measuring better not move and the way you take point will be very important.  3d just takes into consideration alot of factors. Instead...

I would just set an alignment on the length bar provided.  Table surface z, 2 point for a Line on the side of the length bar for y rotation and a point for x origin.   Probe the features on the length bar and measure distance, select distance in y 

You need to set your alignment correctly.  If your using 3d to measure then whatever your measuring better not move and the way you take point will be very important.  3d just takes into consideration alot of factors. Instead...

I would just set an alignment on the length bar provided.  Table surface z, 2 point for a Line on the side of the length bar for y rotation and a point for x origin.   Probe the features on the length bar and measure distance, select distance in y 

This is the length standard provided with the arm (has the conical seats designed to self-center the master probe).  All clamped down.  RDS reads the length 100% fine, regardless of orientation.

Length bar is calibrated and should provide what the nominals are.  Let's see how off you are.  

turn OFF probe comp, Pcdmis will comp a single point in the direction the probe is pointing.

Will give it a try.  The errors were also noted directly off the probe readout window (and manual calculation), which is ball center, however.

Off by anywhere from 1-3mm, but only in PC-DMIS.  RDS is fine.  Only checked bar due to bizarre readings in conventional program.

Yeah so try measuring the bar using pcdmis. Not in the rds.  

If you aren't using a cad model when probing it's very important how you probe.  The probes/hits you take when creating your features for alignment can be affected and then that affects your measurement.  Use a cad file so you can see where those point are If possible.  And use individual points then create features using those points for alignment.  You'll be able to see if there's any funny business going on

When performing the measurement inside PC-DMIS, you need to turn off probe comp, centre the tip in the cone on the scale bar and record a single point (at ball centre), repeat for the next cone position and then do a 3D distance between the two ball centre points.  It should not require an alignment but you do need to make sure the scale bar is held down sufficiently enough so that it doesn't move.

These are the results of a very quick test I just did myself...

We did.  PC-DMIS still reporting bizarre and inconsistent results.  This is new behavior.  Service ticket was submitted with no response as of now.  Would you recommend re-installation of PC-DMIS? 

A re-install wouldn't hurt - just to rule it out.  Are you using the "center_reference_probe" or a different tip?  The length bar calibration is always done with the centre reference probe.