I'm going to paste this question from an inquiry I just sent my reseller.   Here's hoping somebody here is more familiar or can confirm/deny this condition:  (thanks in advance for any replies!!)

Okay.  Now- to the operational/mate question:
To demonstrate this situation, please create an assembly with three primative parts:
1. A cylindrical donut (to represent a gear).
2. A block (to represent the rack).
3. A rectangle with a hole in it to represent a "gear carrier".

Make the cylinder and the hole in the block concentric and coincident on a surface.  Make the gear rack fixed.  Constrain the block with the hole (now called the "gear carrier") to only allow linear motion paralell to the rack, as in a true mechanical system.

At this point, when I apply the rack/pinion mate, what I would expect is that moving the gear carrier relative to the rack would cause the gear to rotate at whatever pitch diameter is defined.  This is not the case.  However, if you fix the gear carrier and move the rack, it will move as expected.  Strangely, if both are free to move, and gear teeth are shown- it is obvious that something is amiss as the gear teeth mesh perfectly when the rack is moved, but move right through each other when the gear or gear carrier is moved with no other variables changed.

Let me know if this problem isnt clear.  Thanks for any help!
-kevin

UPDATE:  I went ahead and built this demo assembly, so go ahead and download this to see what I'm talking about.