Common Datums

Can you share a drawing?  I need to see what the datums are defined as and how they are called out.  It may be that the way it's depicted on the drawing isn't correct.  For example, If A is a plane and B & C are cylinders perpendicular to that plane (think of a face with two dowel holes in it for example), then A controls three degrees of freedom (two rotations and one translation), B-C controls the final degree of rotation about the plane and the remaining two translations, making the addition of C as a tertiary redundant.

I can't post the print but I can tell you the plane and two dowels are all facing the same axis. In a legacy alignment I would level to the plane, square to the two dowels and zero on one of them. Is my understanding of the common datum function correct?

Datum reference frames function differently to traditional alignments.  They behave like a hard gauge and follow datum precedence rules.  Each datum must constrain as many degrees of freedom as it is capable of doing, that is why you see an error for this DRF when you try to add C as a tertiary (see my previous explanation).  

The only way you could add C would be to select "customized DRF" and override the natural constraints but that would be going against what is defined in the standard.

As always, you should seek clarification of design intent from your customer / whoever issued the drawing. 

Sorry, had to go and take care of something. You said "making the addition of C as a tertiary redundant". You're saying all one would need is |A|B-C| to construct that alignment? How are you defining which of the dowels is your origin?

Sorry, had to go and take care of something. You said "making the addition of C as a tertiary redundant". You're saying all one would need is |A|B-C| to construct that alignment? How are you defining which of the dowels is your origin?

B-C is treated as a pattern so It puts the origin mid-way between them - the average centroid of the pattern.

OK. So if you need it to report from origin-x, you would have to make an alignment and select to report from the current alignment rather than the datum reference frame. Dimensions will then match print, but the results will come from the datum reference frame position. Do I have that correct?

Yes