Vijay Sankaran posted to TechNet a plea for solder acceptance requirements 
     to be based on science rather than opinion so that the cost of needless 
     rework could be avoided (my synopsis).
     
     I think that he is on to something and I hope that the discussion pro- 
     ceeds to a good conclusion. The relationship between design, service 
     environment, and inspection is one that deserves to be considered 
     carefully.  I had been studying solder connections for many years 
     before I came to realize that most connection failures I could find 
     documented were caused by conditions that could not be detected by 
     inspection.  These included solder in tension, faulty plating (brittle 
     copper and inclusions), and stress relief that was inadequate by 
     design.  I am concerned that most of the connections being reworked 
     today are not at risk of failing, and that the connections that get 
     selected for rework depend mainly on who looks at them.  This is 
     obviously wasteful.
     
     Given the difficulties that have been documented in getting inspectors 
     to agree on their calls for solder connection nonconformances it cer- 
     tainly is unrealistic to expect operators and inspectors to be able to 
     judge a connection for reliability without something to go on besides 
     001 and 610.  The way I see it, the designer has two responsibilities: 
     1) using the known service environment, to identify scientifically 
     sound acceptance requirements for the various kinds of connections, 
     and 2) to ensure that a properly prepared connection is reliable 
     (plating and finishes properly specified, adequate mechanical com- 
     pliance of connection members, solder not in tension).
     
     That said, for the most part, the designer will not be able to do item 
     1 because he/she does not have the necessary training.  What could be 
     done is for people who have specialized in solder connection reliabil- 
     ity to identify certain classes of service environment and then pro- 
     vide acceptance requirements for each.  (The person I know of who has 
     done the most along this line is Werner Engelmaier.)  What would come 
     out of this study (a combination of analysis and experimentation) 
     would be, for each type of connection, such as through-hole connec- 
     tions, gull-wing connections, connections on a 28 I/O leadless chip 
     carrier, etc., the minimum requirements for reasonable reliability for 
     each class of service environment (not the existing Class 1, 2 and 3 
     of J-STD-001).
     
     For connections that are properly designed and specified, another way 
     to look at the question of how to inspect for reliability is for the 
     factory to control incoming materials and manufacturing processes well 
     enough that abnormal connections are rare.  In such a circumstance, 
     rework of the occasional abnormal connection is no big deal.  This 
     seems to be the way that things are moving, and it is the way that 
     J-STD-001 is designed.  I have been told that the rate of solder 
     nonconformances is below 100 per million connections at more than one 
     factory. (I failed to find out whether this was machine soldering or 
     manual soldering.)  To take Vijay's example of 60 percent (or worse) 
     wetting, if solderability and soldering are so controlled that such 
     cases occur once a week or once a month, then there is very little 
     monetary impact of reworking those that occur, whether they are at 
     risk or not, and regardless of what class the drawing specified.
     
     One benefit of this approach is that it is much more satisfying to the 
     customer, who may very well not be swayed by pure logic and data, no 
     matter how correct it may be.  As in the old saying, "The customer is 
     always be right".  The risk of a dissatisfied customer may be too 
     great to attempt to get him/her to believe that those funny-looking 
     connections really aren't going to fail, especially when their occur- 
     rence suggests that they are due to failure to control materials and 
     processes.
     
     Gordon Davy

***************************************************************************
* TechNet mail list is provided as a service by IPC using SmartList v3.05 *
***************************************************************************
* To unsubscribe from this list at any time, send a message to:           *
* [log in to unmask] with <subject: unsubscribe> and no text.        *
***************************************************************************
* If you are having a problem with the IPC TechNet forum please contact   *
* Dmitriy Sklyar at 847-509-9700 ext. 311 or email at [log in to unmask]      *
***************************************************************************