Dendrites can cause failure of printed circuit board assemblies. Dendrites can form on printed circuits, on components underneath solder mask or under conformal coating.
Hello my name is Bob Willis and welcome to Defect of the Month. Every month I create a video guide featuring one particular type of defect hopefully explain how it’s occurred and possibly how to overcome it. We hope you find these resources useful as you build up your own knowledge base. Please pass these links to other engineers and subscribe to WNIE TV channel on YouTube.
Dendrites can form on the printed circuit board surface where metalized tracks are exposed. Let me illustrate this with a simple animation. Here we have two tracks, and we have placed a droplet of deionized water on the surface between the two tracks. We’re passing a voltage between the two contacts and what you’re looking at is a little Christmas tree growing, it’s what I call the fern like formation as I tend to refer to it, growing between those two surfaces, this is effectively dendrite formation shown below
The very first time I ever experienced dendrite failures was on a product which was in operation, but only for a few days, but in a very hostile environment. This image below shows how corrosion has occurred underneath the solder mask. Now this really was poor design. The solder mask had been defined right up to the pin/pad which was isolated on the board. All the pins were wave soldered using water soluble flux. The flux had penetrated up as it would do in a normal process, but the flux remains trapped between the solder mask and the pin and the isolated circuitry under the solder mask. When a voltage was applied, dendrites and corrosion products were formed leading to intermittent then burn out of the circuit
Now you can see I have removed the mask from the surface of the board on the second image above. You can now easily see the debris left behind. I had to demonstrate how this occurred, why it could have occurred, and how we could overcome it. The engineering manager was convinced his process was providing satisfactory results with very low contamination measurement results after cleaning and I agreed the fault was down to the design. The ionic contamination meter provided satisfactory readings but most likely down to the solder mask cloaking the pad and pin area
For testing purposes I took a number of printed circuit boards, some I soldered it with water soluble flux, some with traditional low activity flux and some samples I did not use any flux. After testing at high temperature, high humidity with a voltage applied to the test circuits we could quickly see the difference between the reference samples and replicating the failures. It proved what had actually occurred with the field failure.
Now dendrites can occur underneath components, they can occur underneath conformal coating. In the example above I show you how corrosion products have actually been visible or found to be visible on a single sided board. This was caused by poor cleaning. What happened is someone had manually cleaned the board assembly on the opposite side and residue had run through the holes to the opposite side. During operation of this product it failed with intermittent faults. You can see the dendrite is very, very obvious.
Now, if you’re going to attempt to look and try and find any dendrites which may be causing intermittent failure of a product, you’ve got to define roughly where it is, you really want to define using test engineering support to say it’s around that component, underneath that termination or it’s in that location to avoid wasting time and removing components before the failure is correctly analysed
When you’ve done that, you can use traditional optical inspection without losing any of the information or displacing valuable information. It’s also possible to do this with X-ray, depending on the size and density of the corrosion products that are left behind. In this example I show dendrites formed under a QFN package just away from the area where solder paste has reflowed.
A final technique we have used in the past is microsectioning. Now, this does sound a bit strange sectioning through what effectively is thinner than human hair. No what I’m doing is actually sectioning through the board up through the bulk/thickness of the board assembly. So I’m coming up underneath where the component terminations are and I can then see through the epoxy resin, pads, solder mask then see dendrite formation. You can see this in these optical photographs.
Now if you want to test materials to see that they will meet your requirements there are IPC standards for assessment of Surface Insulation Resistance SIR. We can use high temperature and high humidity environment; these traditional tests have been around for many years and are very effective.
However, we can now take it on to the next level with far more demanding tests that were developed by NPL in Teddington in London, this is condensation testing. We take printed circuit board assemblies, test boards like these examples which I have built many times for customers. These all have test patterns directly underneath the components below which allows us to assess residues that might be left behind because you’re running a no clean process. Or if you are cleaning, it demonstrates whether you’ve been successful in removing all the material from underneath the packages by measuring the SIR or changes in SIR measurement between a clean and a no clean and perhaps a reference board that has not been subjected to any other operations.
Condensation testing condense moisture onto the surface of the PCB assembly. The photographs below, which are actually taken from a time lapse video, shows how the condensation is slowly increasing on the surface in the two images.
The test method and results and based on the extensive work NPL, they can define how much condensation is taking place on the surface, so that then allows the team to estimate the difference in reliability of products which have gone through a process with a particular type of flux, using some other metallisation or some other process profile on the printed circuit board. It also allows you to look at the sensitivity of smaller and smaller conductor spacing in future projects.
We have many Defect of The Month videos and WNIE Online articles which we hope will help you solve your process and product failures. We have listed a small selection from over 100 plus videos below created over many years with NPL, IPC & WNIE
BGA popcorning
Dendrite formation
Open solder joints
Solder skips
Coating bubbles
Ultrasonic damage
Missing components
Incomplete past print
Component cracking
Tombstone chips
Solder balls
Solder bead formation
Solder shorts
Sulphur corrosion
Crimp connection failures
The list goes on and on, one month at a time
If you would like to download any of our Process Defect Guide booklets or my three soldering text books FREE just drop me an email bob@bobwillis.co.uk
