Andy-

I looked through a few recent, and older, papers this morning, and learned a
few things:

1) I believe the answer to your question depends, roughly, on package
type/style (size?).
What plastic packages are you concerned about? PQFP? CSP? TSOP? PBGA?

2) Much has been written about PBGAS, especially by the people at Motorola.
Some papers I have listed in a "references" file include:

a) Diffusion Model to Derate Moisture Sensitive Surface Mount ICs for
Factory Use Conditions ... "derates JESSD22-A112 and IPC-SM-786A moisture
levels for factory conditions other than 30°C/60%RH...
IEEE Trans., CPMT-C, April 1996, p110
(10) 12 ref
Shook, Richard L., Conrad, Timothy R., Sastry, V. Srirama, & Steele, David
B. [all AT&T Bell Labs]

b) Moisture Uptake and Release by Plastic Mold Compounds- Its Relationship
to System Life and Failure Mode
IRPS Proc., April 1980, p293
(8) 2 ref
Dycus, Dale W. [Mostek Corp.]

c) Effect of Mold Compound Components on Moisture-induced Degradation of
Gold-Aluminum Bonds in Epoxy-encapsulated Devices
IRPS Proc., April 1990, p244
(7) 11 ref
Gallo, Anthony A. [Dexter Elec. Matl. Div.]

d) The relationship between moisture resistance and epoxy molding compounds
in integrated circuits
J. Electronic Matl., 1984, C- p231
(20) ref.
O. Nakagawa, I. Sasaki, H. Hamamura, & T. Banjo [Mitsubishi Elec Corp]

e) The PBGA: A systematic study of moisture resistance [Best Paper]
Proc. IEPS Conf. 9/94, September 1994, C-p588
(15)
Hawkins, George, Ganesan, G., Lewis, G., & Berg, H. [all Motorola]



3) Guessing that you have a PBGA package, I'd highly recommend you see paper
2e) above. There's a wealth of information in this paper regarding the
materials and interfaces that may be the source of popcorn susceptibility.
It's amazing that Motorola sorted all this out in 1994!

The IEPS paper says...

"...Since failure initiates in the die attach zone, it is the moisture
entering the die attach that is most critical." [page 590]

Figure 2 of the paper is a plot which shows, roughly, popcorn susceptibility
when % weight gain <due to moisture> is greater than 0.23% @ 74% RH, and
0.17% @ 53% RH

Figure 7 of the paper states that for a particular PBGA studied, the
"critical moisture concentration in die bond = 2.475E-02 grams/cubic cm"


4) It appears there is no simple answer to your question, based on my
skimming through paper 2e). You could probably say that for a given type of
PBGA,

x by Y mm in size, with xxx balls, full array, "average" die size = xx x yy
mm, using "typical" overmold epoxy compound X, die attach material Y (Ag
filled epoxy), with PWB substrate [BT material, XYX soldermask over Au/Ni/Cu
die attach pad, pad metal exposed only minimally to die attach material.],

popcorn susceptibility for your reflow process occurs ABOVE (0.2??) % weight
gain due to moisture absorption. 



BOTTOM LINE: This probably yields a decent rule of thumb limit for
acceptable moisture absorption, which can be calibrated to ambient air
exposure (RH= yyy%) for xxx hours in your particular assembly floor.

Hope this starts to help.

Michael Alderete



Andy Mackie wrote...
------------------------------

Date:    Thu, 22 Apr 1999 15:33:48 -0400
From:    Andy Mackie <[log in to unmask]>
Subject: Popcorning

Hey, Techdudes: Quick one for you....

How high does the (water vapor) pressure in a package have to get, in order
to see popcorning occurring? I know it will depend on the package etc, but
pitcher's mound (i.e. "closer than ballpark") figures would be great.

Thanks!

Andy

------------------------------