definitive footprints

Has anyone noticed how SMD reflow mounting footprints for a given package can vary greatly from manufacturer to manufacturer? I went thru my libraries and found a half dozen different footprints for sot23. I guess it doesnt matter a whole hell of a lot, but aren't these things supposed to be standardized?!

I'm having a bit of a problem with the sot363 package in general. my surface mount contractor is telling me my footprint in sub optimal for the micr0 c0mmercial quad-diode in this package, but I just checked it and it is dead-on with what NXP published. it has worked for me in the past with other parts in this package. any input greatly appreciated.

excerpted from a nice white paper I found on AVX's website:

"Pad Design. Solder fillet shape and size is ultimately
determined by the pad design. Unfortunately there is a
profusion of recommended pads by component manufacturers,
government agencies, industry associations,
and component users to name a few; but which has
“THE PADS” is the question. Pads recommended by
government agencies result in bulbous, easy-to-see
solder joints that are neither reliable nor manufacturable.
Excessive solder increases soldering defects
(tombstoning) and makes components more sensitive to
handling damage (cracks). Remember when you sit
down for an IRS audit and the agent says, “I’m with the
government and I’m here to help you.” Would you
believe him?"

See:
http://www.avx.com/docs/techinfo/smzero.pdf


IPC publishes at least 3 sets of footprints for each package. yeesh.

mikep

Can't really help you here, other than to empathize.

But that is a great white paper quote! :thumb:

In general I'd ask the CM to be specific about what their problem is with the pads as they are the ones who must make it work.. If you don't trust the CM's advice you may need a different CM. They should know what they can handle with their process.

That said not all CMs are top notch... I had one fry some sensitive film caps I tried to use, that according to spec should have survived their process, but didn't. And this was before ROHS which may be higher temps yet.

JR

good point, John, but I think our CM is quite good actually. I have specific information from him that isnt really crucial to the conversation, and they are making it work. We have a direct feedback path that turns up non-critical issues that may never make it back to the source in other less personal b-2-b relationships. Since I am involved in hi-rel mission critical yadda yadda type work almost exclusively, I gotta look into it, try to figure out if there is a "better" way to do it next time. turns out this is quite a can 'o worms!

Mike P

Pad design is a bit of an art form. Like you said, you would think that footprints would be standardized. But they're not. Without being a dickhead, I think most datasheets show the "recommended" pad design. That is what that manufacturer is recommending. You may have to tailor your actual pad design to fit the process as we have already mentioned here. Just recently we spent a couple of months (not full-time) trying to get the soldering correct on a very small DFN package with our CM. They finally found out what the problem was and now those DFN's are getting soldered reliably. We had to re-layout the board to modify the pads and the solder mask apertures. It turned out that that wasn't actually the problem but it did help get us moving in the right direction. DW.