Lead free BGA's & std SnPb paste| 30 October, 2003
Now I've checked the archives on this and found the starts of an answer but not much else, but then that posting was a while ago and I though there would be some more data now. So here's the question: who, if anyone is placeign Pb-free BGA's (ideally with the Sn, Ag, C allow ball) on std Sn/Pb paste and have you any issues ?
My thought process so far:
For the ball melting point we're looking at 217 deg C,
Most of the work on Sn/Pb & Pb free has been designed mixing Pb free paste with Pb products. Generally 0.5% Pb contamination is enough to create an issue and reduce the reliability of the end joint, some of the Nemi work supports this http://www.nemi.org/newsroom/Articles/CA-Pb-free0602.pdf In addition the intermetalic growth rate of the Pb free alloy is greater so control of the cooling is important.
Motorola however has done some work on SAC alloy BGA balls and SnPb solder and haven't found issues( http://www.mot.co.jp/about/ehs/green/ENV_PBFREEOV.pdf ) as long at the temperatures are high enough to actually get the ball to melt and recommend it's not below 225 deg C. This however will actually push you into going out with the spec's of the rest of the components on the PCB, I would suspect, as most of the components will be capped at 225 deg C due to IPC MSD limits, some of the devices may be suitable for 240 deg C based on the jedec MSD stds but it's best to work on the assumption that they are capped at 225 deg C Voiding in the std BGA devices if run above the 225 spec limits will be fairly high and outwith the IPC limits, there is also likly to be hi voiding int he Pb-free device. What of the long term reliability?
Other thoughts, what happens if you keep the temperature low, say 208 deg C, you then turn the Pb-free device into a high melt device .... however will I get the intermetalics I need for the long term join and what will the wetting be like ?, gut feel is that it'll be slower so shallower again long term issues?