All, thanks for the great responses. Maybe I wasn't as clear as I should have been. I was not talking about the demension of the fence to the mounting plate, but the fence to the table top. It looks like a solid, precision piece, but it is not. The mounting plate is as level as it can be. Not only did I drive myself nuts with a straight edge, but I also ran different pieces of wood along it to make sure the boards didn't hang or jump as it slid onto, and off of the mounting plate. The measurements I took with a square (when I discovered bad cross sectional cuts) was at the juncytuion of the fence and the aluminum table top. According to the Bosch tech rep, the table top will have a/ up to .040" convex, or crown to it. Here, here is the reply from these people:
From: Robert Bosch Tool <boschtools@speedymail.com>
>Reply-To: boschtools@speedymail.com
>To: hwat44@hotmail.com
>Subject: RE: Bosch Tools Contact Us Form [#1089446]
>Date: Tue, 30 May 2006 11:32:10 +0000 (GMT)
>
>Dear Randall,
>
>Thank you for your patience. The RA1180 table top is crowned, it is not
>perfectly flat. So, if you're trying to achieve a square from the fence to
>any point on the table top, it's not going to happen.
>
>Here is the official explanation from product management:
>
>None of the tops are dead flat, but that is by design and if the tolerances
>are correct the table works fine.
>
>Regarding your inquiry relative to the flatness of the Y2K router table
>top, the top is to be within .040" flatness overall, preferably convex.
>However, the .040" could be entirely concave and still be within
>specification.? In addition, there are many combinations and possibilities
>in between the maximum limits, i.e., .030" vertical and
>.010" under nominal.? If the table is evaluated with a straight edge, the
>measurements must always be from any outside edge to the center.? If the
>table is crowned .040" in the center the measurement is only to be
>taken to the center.? If the measurement is taken across the diagonal or
>side to side and the table is crowned, the measurement observed must be
>divided by 2.? If the table is concave then any reading would be measured
>to the center and be the measurement observed.