dendritic growth

We have a problem with dendritic growth that is affecting a new board going into production. This is of concern because they were subjected to Damp Heat Cyclic Tests as part of the product validation testing during their development.

Fails to date

1.Fast track boards, populated and varnished. Unacceptable fail rate. Severe crystal and dendritic growth causing shorts. Oven profile checked and found to be good. Trials done at higher than recommended profile temperatures to ensure flux had been activated. Made no difference.

2.Fast track boards, cleaned then populated, and varnished. Unacceptable fail rate. Obvious crystal and some dendritic growth causing shorts.

3.Fast track boards, populated then cleaned and varnished. All passed. Little evidence of crystal growth. No evidence of dendritic growth.

Results suggest that starting off with clean boards has an effect. The boards may be dirtier than normal supply by virtue they were from a fast track supplier. However starting out with clean boards, whilst it reduces the problem, is not the solution.

Results show that cleaning the boards after population solves the problem and suggests that the solder paste and the �no clean� fluxes therein significantly contribute to the problem. However introducing cleaning in the process is not considered to be a particularly attractive solution.

Understood that lead free pastes are less prone to crystal and dendritic growth under these conditions. Also advised that non lead free components can be soldered using lead free pastes. (NB have since had conflicting information � this needs to be further checked out)

Clearly we do not want to clean after placement as this is very painful in high volume production.

Has anyone any ideas what else this might be or a better way round this problem?

reply »