If I had a nickel for every time my QA dept asked me this same question......

The spec (believe it or) is on Surface Mount Land Patterns (I think SM-782 ???) this has some data in it some where in it you will find 1 min - 2 max pages of info on the subject.  

The reason the boards are 'larger' is that you are effectively breaking glass fiber bundles, this causes a jagged, 'fractured edge' (at the .015" remaining thickness = 'web'.  Their is no way around this 'fractured edge effect' due to the fact that every time you break a score line - you get it, in one form or another.

My method, to remove, minimize, or get a 'true board size'.

Option 'A' - Easiest
1) Use the pointed end of your calipers - this end can penetrate the fractured area and give you a more accurate, true board size. 

Option 'B' - More difficult
1) Obtain some the throw away rail pieces from your panel/array/pallet (whatever you call it)
2) Use one of the scored edges from this scrap as a sanding board to remove the burr from the PCB's at a 90 degree angle in the 'z' move both 'x' & 'y'
3) I have found that this easily removes .005"+, which does bring you to typical IPC specs for routed edges.

Option 'C' - Most difficult
1) Tape down a piece of sand paper on a convenient table (about 440 grit)
2) Stack PCB's flush.
3) Run a manageable stack across this at a 90 degree in the 'z', steady 'x', move the 'y'
4) Excessive use of this method can reduce the board size below spec if not done carefully.


(Below this line is where I get the dope slap......)

In my opinion, the spec is not very informative, nor accurate in content ......IN MY OPINION.

Get a copy of the IPC EXPO 1997 proceedings S-10-5, in that section, you find a paper that just so happens to be on this subject.  It does have attainable tolerances quoted etc...(by the way I co-wrote the  paper with Ron Murdock of NLS Tech).  Also as an added bonus, in S-10-6, you will find another paper on scoring by Mark Simmons of V-score Central.

Joseph E. J. Duclos Jr.
ECC Corporation
Street address: 156 Princeton Street, Holden, MA 01520
Mailing address:        P.O. Box 176, Jefferson, MA, 01522
E-mail: [log in to unmask] or [log in to unmask]

DISCLAIMER: The above statements are my own, and not necessarily that of my employer.
(Don't you hate the legal stuff)


-----Original Message-----
From:   Padmanabha Anandapuram Halappa [SMTP:[log in to unmask]]
Sent:   Friday, January 16, 1998 12:39 PM
To:     [log in to unmask]
Subject:        [TN] BURR

     Good day Folks,

     We face dimension problem after depanellise the v-scoring board. The
     board's thickness is 0.062" and the v-scoring remaining thickness is
     0.015". After depanellise, the board dimension is found to increase by
     0.01".

     According to the fab. house, this is caused by the burr (protrusion)
     after depanelling the board.

     Is there any standard spec. for the burr after depanelling the
     v-scoring board? What is the general dimension tolerance used in
     design the board with v-scoring?


     Thanks


     Robert