Localized extraction clearly shows residue volumes at isolated spots.

Ed.: This continues a discussion begun last month.

We can define cleanliness from an observational or theoretical stance, but what do the data say? This is not to say a general understanding of the overall board is wrong, but rather that we need to understand the cleanliness of the critical areas. Using results from a large investigation of process variables, let’s look at two process variables on Umpire Test coupons.


Group 1 used dirty (high chloride and bromide residues from HASL) incoming boards processed with water-soluble solder paste and wave flux, then cleaned with DI water. Group 2 used dirty (high chloride and bromide residues from HASL) incoming boards processed with water-soluble paste and wave flux, then cleaned with DI water with 3% saponifier added to the wash solution.

The testing plan included localized extraction of the 80-pin TQFP using the C3 extraction system; then the 68-pin LCC breakaway coupon was ROSE-tested and the total board was bag extracted for ion chromatography and modified ROSE testing.

Group 1 (No. 26)    Cl    Br    WOA    OM600    SMD
Bag extraction (total)      5.1    7.24    5.04    -
ROSE testing (LCC)    -    -     -    3.7 µg/in2    NaCl equiv
Modified ROSE    -    -    -    9.3 µg/in2    NaCl equiv
80-pin TQFP leads    9.51    14.57    11.43            -
(C3 results 41 sec. = dirty)

Group 2 (No. 72)    Ch    Br    WOA    OM600    SMD
Bag Extraction (total)    1.05    3.22    1.53    -   
ROSE testing (LCC)    -    -     -    3.0 µg/in2 NaCl equiv
Modified ROSE    -    -    -    8.1 µg/in2 NaCl equiv
80-pin TQFP leads    2.65    0.57    4.40            -
(C3 results: 151 sec. = clean)

Tables 1 and 2 show the Umpire Assembly SIR results for both groups.


The results reflect only a single board from each group of a much larger experiment. These individual boards are a good representation of the results of each group. Saponified cleaning (using low pressure) of the water-soluble flux on a dirty incoming bare board does a good job of removing ionic residues from the assembly.

Conclusions

The difference between bag and localized extraction of the boards showed a clear difference between the samples. The sample from Group 1 bag extraction shows acceptable chloride levels as defined by Foresite’s recommended limits (less than 6.0 µg/in2) after cleaning, but the localized level of contamination is high for chloride and bromide, with poor C3 electrical response (which correlates to assembly SIR performance). The comparison between Group 1 and Group 2 reveals the same issues: Bag extraction of the total board shows a lower ionic level, while localized extraction shows a pocket of ionic residue that, in this case, is acceptable and shows good electrical performance under C3 and SIR testing. But comparing the differences between Groups 1 and 2 for the localized residue pockets gives a better residue understanding than do the total board in a bag or ROSE and modified ROSE when compared to SIR performance. A localized extraction gives a clear view of the residue volume between pads on a critical circuit, while the general assembly cleanliness is a different data point, revealing gross surface residues over the entire board, a reflection of process conditions.

Terry Munson is with Foresite Inc. (residues.com); tm_foresite@residues.com. His column appears monthly.