Wait....Stop! Let's let the originator answer the following questions:
     
     1) Is the outerlayer process "Print and Etch" or "Pattern Plate"?
        (Are the holes tented or exposed (solder plated) during the etching
         process?)
     
     2) Is board(s) doublesided or MLB?
     
     3) Do you see any trend or pattern to the location??
        (ie, hole size, board location...etc.?)
     
     Groovy
     
     I see a lot of valid potential sources but I don't think we're all
     talking the same type of processing/flow.
     
     Enlighten us!

______________________________ Reply Separator _________________________________
Subject: Re[2]: Plated Through Hole Rim Voids
Author:  [log in to unmask] at SMTPLINK-HADCO
Date:    12/20/95 11:43 AM


     Could it be outer layer image misregistration resulting in one side of 
     the hole not being covered by resist?  I've seen something like this 
     with severe cases of outer layer misregistration so that there was no 
     annular ring on the outer layer on a via.  My theory being some 
     etchant gets in the rim of the barrell on one side.
     
     [log in to unmask]
     
_____________________________ Reply Separator _________________________________ 
Subject: Re: Plated Through Hole Rim Voids
Author:  [log in to unmask] at Dell_UNIX 
Date:    12/20/95 2:19 AM
     
     
The question is, is the void caused due to lack of sufficient copper 
coverage before etching, or, is the void due to a lack of etch resist 
coverage during etch?
     
I have seen this before in the "old days" when solder plating used for etch 
resist was out of control. These days, due to HASL finish there is little 
attention generally given to etch resist plating coverage... that is until 
you have a problem. What is your etch resist method? Is the chemistry in 
control? Taken any microsections before etching lately? Check for organics? 
How are your anodes?
     
Hope this helps... good luck
Dave
     
     
>We have experienced a random problem with three or four parts relative to rim 
>voids.
>
>In some cases it was part of the lot and in others it was the whole lot. 
>
>They have been partial, full, top and or bottom. (no consistency) 
>
>They are typically one to two mils down from the surface copper and one to 
>three mils in height. 
>
>They are not specific to any hole size. 
>
>We use Dupont APG715 Dry Film, and develop on a Atotech/Chemcut horizontal 
>system using Potassium Carbonate.
>
>
>
>I would appreciate any suggestions, as to the cause. 
>
>
>
>
>