418 error

[[Pa...]]

We just had an error I've not seen before.  The CMM was running a well-proven program through lunch.  When we returned the stylus was in contact with the part, traffic light was red.  The error was "418: problem with sliding average".

I cleared the error and picked up where it left off.  The stylus made contact with the piece, the sensor (VAST Gold) gave the green light indicating contact, and... seven entire seconds later the machine registered the hit.  The machine moves normally and seems to be measuring fine, but every single-point-strategy cylinder in this program suffers from the same problem; it takes between 1 and 7 seconds to register a hit after contact.  It appears this only happens with this program as I was able to successfully re-run a piece from another lot of a different part number with none of the weird behavior.  There have been no collisions.  The machine has run 2 shifts 5 days a week since calibration two weeks ago.  It's an Accura2 with VAST Gold.  I saw planes run, both single points and scanned; no problem.  I saw 2D curves run without any issue.  I changed the affected cylinders to other stylus systems and the problem persisted.

I ran a qualification (Master Probe and another stylus), no issue.

All cylinders mentioned here are oriented along the Y axis and at very similar X/Z positions.  Probing forces and speeds are identical.

It's an obscure problem.  Is it familiar to anyone?