| Hi Everyone, | We plan to build, for reliability/experimental/research purpose some CSPs (namely Tessera microBGA 46-I/O 0/75mm pitch and TI 64 I/O 0/8 mm pitch). the various factors that we plan to evaluate are the reliability and assembly concerns for four diff pad types: full microvia in pad, partial via in pad, the dog bone and the normal circular pad design. | We also plan to play around with the permissible placement offsets that these pads can take with allowable self centering. | I would really appreciate if anyone of u cud tell me the DOs and DONTs before we embark on the experimental design. Also any type of trends that u all have found for for similar experiments. | Thanks in anticipation | Regards | | | Parvez Patel | Graduate Research Associate | Universal Instruments Corporation | Binghamton-NY 13905 | Ph: 607-779-7215 | | Parvez,
I've been down, and continue, this path many times. In 1982, I created standard hole in pad configurations using sequential lamination and plating over the remaining hole in pad structure so no hole was apparent. This was a very reliable situation because it relied on proven technology.
In 1993, I used standard, high aspect ratio, plated through holes (10:1 as 6 mil holes in .062" thick MLB's) in BGA pads and as blind vias in same. Of course, the through holes didn't work well because of solder drain off. The blind vias worked well but some outgassing was apparent - depending on flux, paste, and profiles.
The problems I'm finding with micro-vias is similar as far as some outgassing - depending on depth through the outers. The biggest problems concern fabrication relative to hole wall preparation (etch back, etc.), electroless copper deposition (very often inconsistet), and electroplated copper often low on ductillity and also inconsistent.
I have much more, but no time here to say it.
Have a grand experiment,
Earl Moon
reply »