Minimizing Tombstoning Defects

Paste Flux / Reflow: Our experiments indicated that the combination of no-clean solder paste and air forced convection reflow produced the fewest tombstone defects, followed by water-soluble paste reflowed in air, and most tombstones occurring with no-clean paste being reflowed in a highly concentrated nitrogen atmosphere where average oxygen levels were measured to be less than 50ppm. Further investigation showed reduced tombstone defects using the no-clean solder paste in a reflow atmosphere with lower nitrogen levels. Under nitrogen rich reflow conditions (i.e. < 200ppm oxygen), the trend observed was larger dissimilarity in the tombstoning levels produced by different vendors no-clean solder paste products.

Paste Alloy: It was noted that the Sn/Ag/Cu no-clean solder paste produced fewer tombstone defects in a very highly concentrated nitrogen forced convection reflow atmosphere (i.e. 20ppm oxygen). The thought is that the slower wetting speed of the Pb-free paste is advantageous under this scenario, providing better balance between the surface tension forces of the solder on the component terminations in reflow. The result is the Sn/Ag/Cu solder achieving a more stable reflow process (under high nitrogen reflow) than the faster wetting action eutectic Sn/Pb solder.

Pad Dimensions: The pad design producing the fewest tombstone defects out of 27 tested was 15 mil pad length, 12 mil pad width, and 9 mil pad separation. The corresponding aperture design for this pad was 15 mils long x 11 mils wide (rectangular openings) utilizing a 5 mil thick laser-cut stencil.

Amount Of 0201 Component Overlap On Solder Paste Deposits: With perfect print registration and high placement accuracy, this variable was not found to be significant on tombstone defects. With any misregistration of print or placement, there will be an unequal amount of component overlap across component terminations. This situation could potentially upset the balance in wetting forces across terminations. Tombstone defects were found to be most prevalent when intentional misregistration was introduced under a low designed overlap scenario. The highest overlap tested (i.e. 6 mils) was found to produce the fewest tombstone defects with intentional misregistration. The trade-off with using a high overlap is that there is higher potential for trapping solder under the component and forming solder balls and for tightly spaced components there is a greater danger of solder paste bridging defects as solder may be squeezed out under placement pressure.

Component Reflow Orientation: The conclusion of this analysis is not so straight forward, as results depended on the process conditions under test. This variable was not found to be significant for assembly with no-clean paste and forced convection reflow in air. However, for builds using no-clean paste and water-soluble paste being reflowed in nitrogen (i.e. < 50ppm oxygen), there was a statistical difference in the number of tombstone defects produced between the two component orientations passing through reflow. Components with terminations aligned end to end (i.e. one end passes into reflow before the other end) produced more tombstones than components aligned side by side (i.e. terminations passing through reflow simultaneously).

-Jeff

This message was posted the OnBoard Forums

reply »