BGA vs QFN in high vibration environment?

[MagicSmoker]

Wanted to solicit the opinions of the eng-tips crowd on this. Got a software engineer friend who is writing the code for some doodad that controls/monitors an IC engine in some gadget (details obfuscated to protect the guilty). He asked me whether a QFN or BGA package of the chip would better withstand high vibration. I thought the BGA - assuming it passed xray inspection - would do better because of the greater solder joint area and smaller distance between said joints, but that was my off-the-cuff opinion and not based at all on fact or even personal experience. So I figured why not complicate matters and ask for more unsubstantiated off-the-cuff opinions?

 

[MacGyverS2000]

Do either, then pot the puppy... about the best vibration resistance you'll get without going to active damping.

I'd go with BGA, though, as the size of the chip increases... less moment of inertia on the pins.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[MagicSmoker]

Potting isn't an option, just conformal coating (polyurethane-based, is my/his understanding).

So, that's one slight nod (in addition to mine) for BGA.

 

[Comcokid]

For QFN or BGA, assuming your board has a rigid mount and is not going to flex, either one is fine. It will be PCB flexure under vibration or resonance that causes problems. That is assuming you don't have some huge BGA package with a half-pound of heatsink and fan hung on the top of it.

Where the issue usually arises is with thermal cycling as there is still some difference in coefficient of expansion between the package and the PCB.