I was interested to read about the two reliability ratings for 
     IPC-SM-840 solder mask that Doug Pauls described in a recent TechNet 
     posting.  The two classes are T - telecommunications (for which a 
     failure is not life-threatening) and H - High Reliability / Military 
     (for which a failure is intolerable).  It got me to wondering about 
     the extent to which the differences in the document (or any other 
     document that has such reliability categories) are based on actual 
     failure rate data.  That is, where a document says something like 
     "This is OK for Class T, but it's not OK for Class H [because it will 
     (or might) reduce the reliability of the equipment]", what is the 
     basis for making the decision?  If it's based on the engineering 
     judgment of the people who voted on the spec, without actual failure 
     rate data to support it, should that be stated somewhere?  If failure 
     rate data were used, should they be referred to in some way (e.g., 
     footnote or Appendix) in the document?
     
     Other examples might be permitted levels of flux residues on a CCA or 
     the amount of, and degree of wetting by, solder in a plated-through 
     hole, as specified in J-STD-001.  Of course, these are just examples.  
     What I'm asking for is some attention to and documentation of the 
     _process_ by which limits in standards are set.
     
     I know that there are people who specialize in risk analysis, 
     reliability, and product safety.  Perhaps some of them are TechNet 
     subscribers and would like to respond to this inquiry.  As I see it, 
     it makes at least as much sense to make reliability distinctions in a 
     standard on the basis of _risk_ of failure as on _consequences_ of 
     failure.
     
     Risk of failure, I believe, is a matter of the service environment.  
     With this line of reasoning, the reliability classes in a standard 
     would be something like:
     
     O - Office (small swings in temperature and relative humidity, low 
     levels of gaseous pollutants like salt and sulfur compounds, no 
     significant vibration or mechanical shock)
     
     G - Ground-based stationary outdoor installation (wider but slow 
     swings in T and RH, still no significant vibration or shock)
     
     Mo - Mobile ground-based installation (possible exposure to industrial 
     and marine atmospheres, higher levels of vibration and shock)
     
     Ma - Marine installation (shipboard or on the ground within a few 
     miles of an ocean, low vibration and shock, high RH and salt level)
     
     A - Airborne (subject to vibration, rapid swings in T and RH)
     
     U - Under hood (large swings in T and RH, high vibration and shock)
     
     Maybe someone has already made such distinctions (I think I remember 
     seeing something like it from Werner Engelmaier), but I don't recall 
     seeing them incorporated into a standard.
     
     In summary, I'm raising questions about:
     
     1. The process by which reliability distinctions get into standards, 
     and
     
     2. The relative merits of making the distinctions on the basis of 
     consequences vs. risk of failure.
     
     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]      *
***************************************************************************