|
|
|
|
|
|
|
||||||||
Review of Tin-Copper and Tin-Nickel Intermetallic Thickness
Board Talk
|
||||||||
TranscriptPhil Welcome to Board Talk with Jim Hall and Phil Zarrow of ITM Consulting, the Assembly Brothers. Today we are coming to you from the ITM elegante ballroom. We are here to discuss process, materials, equipment, you name it. Today we have a reflow question. I have a question considering soldering intermetallics. If the formation of tin copper and tin nickel intermetallics is a function of time and temperature, then what is to be expected in the following scenarios? Components reflow solder to a pre-wired board where the solder is liquid at 215 degrees versus hand soldering the component to a board where the solder is liquid at 215 degrees for 3-5 seconds. What would you expect the different intermetallic layer thickness to be? I have read that 1 - 2.5 micron thickness of tin copper is desirable and a .4 - .8 micron thickness on tin nickel is desirable. If when hand soldering the liquid phase of the solder is so much shorter than the liquid phase inside a reflow oven how does one achieve the desirable intermetallic thickness? Wow, they said a lot there. Jim Yes, they did, P.D. First off I want to complement you on understanding that when you are soldering to an ENIG board or an ENIG PIG board, which has a nickel coating, that your soldering is different and your primary intermetallic is different. You have a tin nickel, rather than tin copper. Many people don't realize it. There is a difference. In general with soldering to copper, and that means OSP, HASL, Immersion Tin, Immersion Silver, and all those where the finish goes away and you end up with a primary tin copper. Those from pretty quick. With hand soldering, typically the extra heat of the iron is adequate to do that. The IPC, which trains you how to make a hand soldering joint, has analyzed that. So those intermetallics form very quickly between copper and tin and that is not a problem. Typically having too long and letting them get too thick has been the issue. But I have to say with all of the new analysis of solder joint reliability the issues intermetallics are much more complex than don't let them get too thick. Particularly with SAC alloys where you have the secondary tin silver intermetallics distributed through the joints. Those are tending to be much more significant. So I would not be concerned. In general you don' want to heat your tin copper solder joints any longer than you have to. I have seen multiple soldering operations without any problem. But the reality is just a lot more data so it is not just that simple as the thickness of the intermetallic. With respect to soldering to nickel such as an ENIG board or an ENIG PIG board, those form much more slowly. The length of time is not of a concern, getting sufficient intermetallic. I have to admit ignorance with hand soldering on an ENIG finish. I never took the IPC hand soldering course that would be whether or not when you are soldering to an ENIG finish or an ENIG PIG, when you are soldering to nickel if you need to hold your soldering iron longer and so forth to establish that nickel and tin intermetallic. I have to beg ignorance on that. Phil, do you have any wisdom to shed on that. Phil Not really. The other thing I might suggest, if it is appropriate subjecting scenarios to reliability testing, including accelerated life testing to see if there is any indication the intermetallic is acceptable that way. Does that make sense to you Jim? Jim I have looked at a lot of published reliability tests of that nature and I have never seen anyone come up and say that the intermetallic was too thick here. I think I saw just recently one was published where they subjected an OSP finish to seven reflow cycles. That did slightly reduce the reliability attributed to intermetallic. But that is a really extreme test. Nothing OSP, we have direct soldering to the copper with seven lead-free cycles to 250. Those were lead-free wave cycles to 250, I don't know what the liquid time was. In most cases I am not seeing anybody say the intermetallic here is too thick and that is why the solder joint failed. Phil If I were subjecting something to seven thermal cycles like that I think I would have a lot of other problems to worry about before I start getting concerned about my intermetallic growth starting with what is wrong with my procedure going down to copper erosion since that was a lead-free scenario. We'll keep reading the studies, there is always more and more coming out. Jim We are sure learning more about intermetallics that we didn't know. We were very naive and happy when we were all just using tin lead and could rely on our forty, fifty years of experience to know that joints would be reliable. Phil Well, you have been listening to Board Talk with Phil and Jim. We hope that we answered the question, or stirred up the solder pot. Bear in mind that Board Talk has not been FDA approved. Possible side effects may range from enlightenment to bewilderment. We want to thank you for listening. And remember whatever you do, please don't solder like my brother. Jim And don't solder like my brother. |
||||||||
Comments
|
||||||||
|
|
|
|
|
|
|