Page 240 - Six Sigma Advanced Tools for Black Belts and Master Black Belts

OTE/SPH
OTE/SPH
Char Count= 0
3:4
JWBK119-15
August 31, 2006
Experimental Procedure and Results 225
Table 15.1 Moisture sensitivity levels.
Level Floor life Moisture soak
1 Unlimited at ≤ 85 C/85%RH 85 C/85%RH 168 hours
◦
◦
◦
◦
2 1 year at ≤ 30 C/60%RH 85 C/60%RH 168 hours
◦
◦
3 1 week at ≤ 30 C/60%RH 30 C/60%RH 192 hours
4 72 hours at ≤ 30 C/60%RH 30 C/60%RH 96 hours
◦
◦
5 24/48 hours at ≤ 30 C/60%RH 30 C/60%RH 48/72 hours
◦
◦
◦
6 6 hours at ≤ 30 C/60%RH 30 C/60%RH 6 hours
◦
and the temperature throughout the soak time. The higher the temperature and the
◦
longer the dwell time over 100 C, the more likely a package is to crack if the mois-
ture level in the package is above the critical level. The study concluded that the
JEDEC-recommended vapor phase and convection reflow temperature precondition-
ing profiles are adequate guidelines to determine the moisture sensitivity of plastic
packages, unless excessively long preheat is applied. This forms the basis for using
JEDEC as the benchmark in our study.
From Table 15.1, it can be seen that there are six levels of moisture soak with three
different temperature--humidity settings. Therefore preconditioning testing of differ-
ent products with different moisture sensitivity levels requires three chambers if they
are to be done at the same time. Otherwise, we will have to complete the precondition-
ing test at one temperature--humidity setting before proceeding to another test with a
different setting. This will greatly increase the cycle time of the test, thus reducing pro-
ductivity. Since moisture soaking takes much longer than the other precondition tests,
it is of great value to see if the equivalent moisture soak times at harsher temperature--
humidity conditions for levels 2--6 can be obtained. Given its potential economic value
in terms of lower capital investment as well as shorter time to market, an investigation
into the rate of moisture absorption on a selected set of devices is conducted. The aim
is to develop a model for moisture gain under various conditions, from which the
equivalent moisture soak time at a harsher condition can be derived. Specifically, a
suitable response related to the weight gain process will be identified and a family
of response functions under pre-set conditions together with a physical model will
be established. Analogous to the basic framework of accelerated testing models, the
moisture soaking acceleration factor, and hence the equivalent moisture soak time,
for a preconditioning test on a plastic package can be derived.
In the following section, we outline the experimental procedure and present a set
of sample results. In Section 15.3 various choices for the moisture soak model are
discussed and a regression analysis is performed to establish the most appropriate
alternative. A set of acceleration factors is generated from the model. Finally, some
discussions and conclusions are given in Section 15.4.

15.2 EXPERIMENTAL PROCEDURE AND RESULTS

The plastic leaded chip carrier (PLCC) package with B24 mold compound was selected
as the primary test specimen for the case study. Three lots of 44-, 68-, and 84-lead (pin)