Solder failures.....

the powerpoint from the Aerospace Corp is interesting, at least in regard to the examples that they present. I've seen a number of them before. The ones showing whiskers growing from the leads of SOIC packages (or smaller?) and SMD components are the ones that get my attention the most, as these are what are in the designs that I deal with. The examples of the whiskers on wire wrap posts or the card rack guides (on the shuttle!) are of less interest as they were already very old technologies when the presentation was created.

For the sake of reliability, I had the impression that military and space applications were still using processes with lead. The problem is getting packages with lead based finishes. My experience in doing work for satellites is that the customer was very conservative and was happy using parts that were long out of production as long as the design was proven. We kept some packages of the dice for an ancient op-amp that we had packaged into a can package by Lockheed everytime we sold a system. Compared to that sort of trouble, getting lead based finishes on the pins of a production IC should be easy.

There is room for debate as to whether going lead-free is worth the benefits, but I still haven't seen signs that electronics reliability has been impacted. My employer has bigger sources of trouble... one example is resistors contaminated by sulfur! There is always the possibility that whiskers could be the cause of some no-fault-found returned parts... but there are lots of potential causes. FWIW, conformal coatings and potting are pretty standard in our products, which are a slight step above automotive grade. I suspect that the low-cost commercial products are less likely to use coatings and therefore will be more affected by whiskers.
 
We had a rash of failures at work that appear to have been related to tin whiskers. Several power supplies failed after being in service for a few months to a year. Close inspection of one showed a slight overgrowth of solder where there shouldn't have been any.

Tin whiskers is not likely what you had. It is microscopic but you appear to have macroscopic issues. It sounds more like corrosion that looks like solder and/or bad mfg.
 
That's very interesting. Do you use a lead-free solder, and do you know the metal content of the solder used. I've been on vacation and will do some searching for whiskers at work. To most they would appear as dust or lint fibers and mostly invisible unless using good light and magnification. :caution:

These units came from a manufacturer. I think I have a photo in one of the root cause analysis documents. I'll see if I can find that.
 
Inconclusive research: Had some time today to research and the results were two microprocessors checked, one metal whisker found. Length 1/2 distance between two pins. Required 10x Bausch and lomb eye loupe with correct light. Both boards tested good. Also these boards were cleaned and dusted six or more months previously using for testing.

Three DB15 male checked, two had metal whiskers. One had 1mm fuzz ball inside top right. Under magnified desk lamp, looked like very fine steel wool. Every time the connector was used, the WhiskerS were compressed. The other had several on the outside left of the shell. Had the appearance of very fine wire strands, required 10x eye loupe.

Removed the whiskers and didn't test at this time. More results as available from uncleaned failed boards.
 
Last edited:
Top