>At 10:31 AM 5/22/96 CDT, "Tippit, Jack" <[log in to unmask]> wrote:
>>
>>We have instances where tin/lead coated pins are inserted into gold plated 
>>sockets, etc.  Components even have gold/tin-lead or gold/tin interfaces, 
>>such as an IC socket with a gold contact with a tin sleeve.  Mil-Std-889B 
>>indicates gold as being incompatable with tin or tin/lead.  Has anyone 
>>looked into this issue and is there a potential for corrosion formation and 
>>associated problems?  Again, I just refereing to mechanical contact ( not 
>>soldering ).

>We definitively traced our PGA/DIP problems to fretting corrosion and most
>likely our PLCC problems also were fretting.  Fretting corrosion of
>tin/tin-lead contact interfaces occur best under low level vibration
>conditions, such as vibration generated by cooling fans.  Under vibration,
>micro-movement of the contact interface occurs.  Every time the contact
>moves, fresh tin is exposed.  This exposed tin oxidizes so when the contact
>wipes back again some of the oxides get trapped at the contact interface.
>When enough of the oxide builds up, the contact resistance increases.  If
>the fretting is severe enough, the oxide breaks all electrical contact and
>an open connection occurs.  Reseating the socketed component temporarily
>fixes the problem.

I was hoping someone else would mention fretting corrosion with respect to 
mating tin-lead to gold.  Rudy Schubert at Bellcore has written several paper 
on his studies of fretting corrosion, and gold/tin-lead contacts are not as 
reliable as gold/gold and tin-lead/tin-lead contacts.