Some edge card connector solder joints were reportedly failing open.  This is an optical image of the connector that showed fractures (red arrows) in the epoxy staking material apparently used to attempt to mitigate the problem.

This is an optical image at intermediate stage of micro-sectioning into the suspect solder joints. The displacements (yellow arrows) are very large relative to what the solder joints could be expected to survive.  Note also that the back face of the connector body is not normal to the PWB as it should be, which is the likely cause of the mechanical loading on the solder joints.  When the subject connector is mated with its mating connector, the loads on the solder joints are exactly as indicated by the displacement vectors shown in this image.

These are SEM images of solder fractures as viewed in a parallel and traverse section of the solder joints.  The evidence that these are creep rupture failures includes (1) the large upward displacement of the connector leads and (2) the fact that the fracture propagated through the bulk solder joint rather than through an intermetallic compound layer or at an interface.

So, why didn’t the epoxy staking material work to prevent these failures?  Because epoxy has a very low mechanical stiffness relative to the connector leads, so in spite of the addition of staking material the majority of the applied load is still carried through the leads to the solder joints.

 

These are chip resistors that were reported to be failing with high resistance.

This is a BSE SEM image of a microsection through the approximate centerline of the resistor.

This solder joint appeared to have failed in a brittle failure mode at the interface between the solder and the nickel barrier plating (i.e. brittle interfacial fracture).

The same issue was found on multiple PCBAs suggesting that this was some sort of systemic failure.

  The analysis results suggested that the resistor failures were most likely caused by mechanical stress (likely in bending) at high strain rate.  Another factor in the failure might have been the PWB cut-out nearby the resistors.

  Other factors to consider are  (1) degree of warpage of the PCBA after reflow and (2) relatively low PWB stiffness which could be improved by increasing the PWB thickness or adding stiffeners for example.


Response to Kurt Larson’s comment … ” It is also likely the reflow process held the solder in liquidous for too long and a too high peak temperature. This would dissolve the solderable layer on the resistor base plating into the greater bulk of the solder fillet. This is a common problem when the temperature of the reflow process is increased to speed up the process, and the process is not profiled for assemblies of varying thermal mass.”

These images suggest that in this case the reflow process was nominal based on the thickness of the nickel diffusion barrier plating.

The image below shows a row of solder joints on a large BGA near the center of the row.

The corner balls on the BGA were elongated into an “hour glass” shape due to the warpage of the package (e.g. image below). 

The chart below shows the solder joint height versus the distance from the corner ball for three different assemblies (Ax, BX, & CX).

The amount of warpage in BGA assembly is driven by a number a factors including the BGA design, the board design, and the reflow process parameters (e.g. top versus bottom-side temperature and cool down rate).  In the present case, it appeared that the BGA design (i.e. CTE constrained by large die) was the primary factor.

This is the solder-side of a PCBA that failed shorted.  The cause of the short was determined to be electrochemical migration (ECM), but there were secondary effects as shown in the next image.

These are vapor deposited copper & tin crystals that most likely formed as a result of the failure where vaporized copper & tin condensed as these feathery structures.

The EDS spectrum suggest the composition of the crystals includes copper & tin.

This is BSE SEM image of a microsection of a flex circuit showing two internal copper conductor layers.

This is an area where the copper conductor is corroded.  The corrosion initially attacks the grain boundaries.

This image shows that the corrosion completely consumed the copper layer at this location(top-right segment).

This is a through-hole solder joint where the internal copper conductors are almost completely gone due to corrosion as well as the nickel surface lands.

This is a higher magnification image that shows the missing material more clearly.

This is an elemental map that also shows the missing copper layers and nickel lands more clearly.

Some tips to avoid this type of corrosion include …

  1.  chose compatible materials (adhesives and other polymers can contain chlorides, sulfur, etc. … add moisture and voltage bias and corrosion  can accelerate dramatically)
  2.  clean thoroughly &  DRY at each appropriate process step (this includes bare flex fabrication and assembly processes)
  3.  thermal damage can create paths for ionic contamination and moisture to migrate into the internal conductors, so minimize thermal exposure

 

This post discusses the differences between optical images and SEM images.

 

 

This is an optical image of a flea.

 

 

 

  

 

 

This is a 20 kV secondary electron (SE) image of the same flea.

 

 

 

 

 

The optical image has good color contrast, which is often complimentary to SEM imaging.  The SEM image has no color (gray scale), but shows far more surface detail and depth of focus compared with the optical image.

 

 

 

 

This is a 20 kV backscattered electron (BSE) image of the flea.

 

 

 

 

The backscattered electron signal is proportional to the average atomic number of the material being imaged, so in this case the carbon tape background and the flea (mostly carbon) are of similar contrast.

 

 

This is a 2 kV SE image of the flea.

 

 

 

 

 

At lower accelerating voltage more of the fines surface details are apparent, but there is less signal and lower contrast.

 

 

 

Higher magnification 2 kV SE image.

 

 

 

 

 

 

 

The advantage of SEM imaging is the high magnification with resolution, which allows for examination of small (often sub-micron) features that would not be possible using optical wavelengths.

FTIR can identify unknown plastic materials by comparison with known materials (i.e. from a spectral library).

 

 

 

Fig. A – Sample A FTIR spectrum compared with Nylon 6/6 and Nylon 6/9. Part of the spectrum
is more similar to Nylon 6/6 and part is more similar to Nylon 6/9.

 

 

 

 

 

  This is the Fourier Transform Infrared (FTIR) spectrum of an unknown plastic (bottom) compared with library spectra of Nylon 6/6 and Nylon 6/9 respectively.  The spectrum suggests that this is likely a blend of Nylon 6/6 and Nylon 6/9.

A client provided four aluminum electrolytic capacitors to SEM Lab, Inc. for analysis.  These were identified as counterfeit by the manufacturer that also examined some of the counterfeit capacitors.

Fig. A – Rubber seal styles for capacitor samples

There were a variety of construction anomalies associated with these samples. These are listed below by sample. 

Sample A – Poorly constructed roll where lead spacing appeared to be bad and close proximity of swage to end of anode foil extremely unusual, as well as proximity of anode & cathode terminations. 5-swage terminations. 

Sample B – same as A (except 3-swage terminations) 

Sample C – 4-swage terminations, position of lead attachment on roll more typical than A or B where attachment was near start of roll. Corrosion damage in beginning stage on anode lead. Seal appears poor and damaged by stave on lead. 

Sample D – 5-swage terminations, position of lead attachment on roll more typical than A or B where attachment was near start of roll. Corrosion damage in beginning stage on anode lead. Seal appears poor and damaged by stave on lead. D foil appears different than C foil.

Fig. B – Sample A – failed capacitor #1. This is the EDS spectrum of corrosion product on the anode near the corroded termination.

The analysis results suggest that the two failed samples became open-circuited due to internal corrosion at the anode lead. These failures may have been accelerated by poor quality seals and construction features that permitted ingress of external contamination (e.g. Cl) and moisture. The not-yet-failed samples also showed initial stages of corrosion failure at the anode lead/seal. 

A client provided a failed metal thin film resistor (Nichrome) to SEM Lab for failure analysis. 

 

This is a failed resistor, IRC P/N W1206R033323B, 332K, 1%, 1/4W.
This is a failed resistor, IRC P/N W1206R033323B, 332K, 1%, 1/4W.
This is an optical image of the microsection at a location just outside of the NiCr thin film resistor pattern.  There is evidence of corrosion at the edge of the resistor pattern (arrows).
This is an optical image of the microsection at a location just outside of the NiCr thin film resistor pattern.  There is evidence of corrosion at the edge of the resistor pattern (arrows).
This is a higher magnification image of the corrosion damage to the resistor pattern.
This is a higher magnification image of the corrosion damage to the resistor pattern.
This region appears to be corroded (i.e. NiCr pattern missing).
This region appears to be corroded (i.e. NiCr pattern missing).
5.    The cover glass and interfacial structure appear to contain micro-pores that might explain moisture ingression from the edge of the resistor.
5.    The cover glass and interfacial structure appear to contain micro-pores that might explain moisture ingression from the edge of the resistor.
This is another image of a region that appears to have been corroded (i.e. missing NiCr).
This is another image of a region that appears to have been corroded (i.e. missing NiCr).

 

The analysis results suggest that the resistor failed due to corrosion of the edge of the NiCr resistor pattern.  The removal of material by the corrosion process created an open (likely multiple opens) in the resistor pattern.