Hi Keith,
I sure wish would be as straightforward and simple, but they are, and were, 
not.
>I have to admit that when I used the word "proven" I was using a more 
limited definition than the one that you >might use.    What I mean is that the 
companies who have been using the alloys I mentioned undertook their >own 
accelerated test program before their adoption and have monitored real time field 
performance since then >to determine whether the confidence established on the 
basis of their preliminary (and in your terms >possibly rather primitive) ALT has 
been justified FOR THEIR PRODUCT by actual experience in service.  
A: If these companies actually performed ALT with regards to solder joint 
reliability, they had little idea of what was needed. And...as Dave has pointed 
out, field failures of product of any kind, and in particular consumer 
products, do not get sufficient analysis do determine failure  ode or root cause.    
>With something like 200 million boards of a wide variety of types produced 
with these solders since 1999 and with a substantial number having now been in 
service for nearly five years there have been, as far as we know, no failures 
in the solder alloy.   
A: While this number sounds impressive, it really is not. First, all of these 
are consumer products--in other words, throw-away products. Second, consumer 
products tend to put very little strains on solder joints because of (a) 
benign use environments, and (b) by the tendency to make consumer products compact 
and thus use small components.    
>There were and still are some failures in connections due to "cold joints", 
which is the same defect that >occurs with tin-lead solder when the joint 
surfaces do not get to the temperature necessary to form a >metallurgical bond 
with the substrate, but that is a process control issue rather than a reflection 
of some >fundamental weakness in the alloy itself.
A: Processing quality issues will always be with us, LF or not.
>I think that to be realistic, Werner, this is the way that the industry will 
have to gradually build up its >confidence in the reliability of lead-free 
solders- do enough "pragmatic" ALT to get sufficient confidence to >use the 
alloy in production and then see how that squares up with field experience- and if 
an
>unanticipated failure occurs go back and adjust the ALT to test for that 
failure mode.
A: This is unacceptable for products were human life is in danger when 
failure occurs. For these 'ALTs' to be at all meaningfull, we have do understand how 
these solder joints behave under cyclic strains in both real applications and 
under ATC.   
>That is the way the industry built up confidence in tin-lead solder. 
Tin-lead solder was adopted in the early >days of the electronics industry when there 
was even less reliability data available than there is now for the >lead-free 
solders. There were some unexpected failures with tin-lead (e.g. due to gold 
embrittlement) but the >industry learned from these experiences and adjusted 
practice accordingly so that overall a high level of >reliability was achieved. 
There is no reason to believe that the same thing will not happen with 
lead-free >solders.
A: No, this is not the way it happened--I was there, remember. It started 
with ceramic components on ceramic hybrid circuits—hardly any strains on the SJs 
at all. Nevertheless, Au-embrittlement caused overstress failures in the 
intermetallic layers---not the solder joints-- on relatively small handling 
stresses. It was only when the industry progressed to polymeric PWBs with larger 
ceramic LLCCCs that all hell broke loose. ALTs and field experience did not solve 
the problem, good research by many workers in the field did that--and it took 
many years. LF-solders are to be substituted into high-density, 
high-technology products that need to function reliably in harsh environments with people's 
lives depending on them--wow!   
>In regard to the current anxiety in the industry that we are moving into 
totally unknown territory, Dr Jenny >Hwang made an interesting and in a way 
reassuring observation in some discussions during Internepcon Tokyo >a few weeks 
ago. She said that so far the behaviour of the lead-free alloys has been 
entirely consistent with >known metallurgical science.    Because metallurgy is only 
incidental to its business of making functioning >circuitry the electronic 
industry tends to have a fairly narrow view of the very large subject of 
>metallurgical science.   While there are around the world many very good 
metallurgists working in the field >of soldering technology the knowledge of most people 
in the practical side of the industry is limited to the >metallurgy of the 
tin-lead system.  People like Dr Hwang, however, have experience in the wider 
field of >metallurgy science and can therefore put the issues associated with 
lead-free solder into context.    The fact >that the behaviour of lead-free 
solders seems to be consistent with the general understanding of the behaviour >of 
metals does not mean that there will be no unanticipated problems but if the 
electronics industry draws on >all the resources of metallurgical knowhow that 
are available it can approach this change with a little more >confidence than 
it seems to have at the moment.
A:  I do not see where you find comfort here.  Saying "the behaviour of the 
lead-free alloys has been entirely consistent with known metallurgical science" 
means just that, the emphasis is on 'known', and unfortunately we do not know 
anywhere as much as we need to.  
>A solder joint does not "know" whether it is in a walkman or a jet fighter.  
 All it "knows" is the mechanically >and thermally induced stresses and 
strains to which it is subject and there can be some overlap in that range >of 
"experience".    
A:  Well, actually a solder joint does "know" whether it is in a walkman or a 
jet fighter, because the strains a SJ is subjected to are quite different 
because of the different use environments and design constraints and thus the SJ 
will respond accordingly. 
>In fact given the intrinsic weakness of tin-lead solder part of the design 
of joints for high performance >electronics has been to try to ensure (e.g. by 
building in stress relief and reducing thermal mismatch) that >the solder 
joints are not subject to stresses and strains that are too much more extreme than 
those >experienced by a soldered joint in a walkman.  
A:  No, there is no "intrinsic weakness of tin-lead solder" and no "stress" 
relief has ever been required for a solder joint. On the other hand, 
strain-reliefs have been required frequently. While this difference may apppear 
academic, it is crucial in assuring that the solder joints function as expected for 
the design life of the product.
>The rapidly accumulating experience in consumer and general electronics can 
provide a  foundation on which >high performance electronics can build. 
A:  There, unfortunately, is no "rapidly accumulating experience in consumer 
and general electronics" that will do that.

Regards,
Werner Engelmaier

-------------------------------------------------------------------------------Leadfee Mail List provided as a service by IPC using LISTSERV 1.8d
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Leadfree
To temporarily stop delivery of Leadree for vacation breaks send: SET Leadfree NOMAIL
Search previous postings at: http://listserv.ipc.org/archives
Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700 ext.5315
-------------------------------------------------------------------------------