One of our customers has experienced a problem on one batch of pcb's we
supplied with immersion gold over nickel.

The ic's and qfp's can be levered off the surface mount pads fairly
easily.  In some instances the component leg pulls out of the solder
fillet leaving solder behind on the pad.  In adjacent areas the solder
pulls off the pad and in some instances the pad pulls off the pcb (a good
joint!).

These variations of performance can often be found on one component.

This customer has reflowed other batches of our pcb's manufactured at the
same time with no problems.

We have checked nickel thickness, phosphorous content and gold thickness.
 All are within specification.  The components appear to have wetted well
in some areas but are pasty in others.  Microsections have not been
conclusive due to the variations in local areas.

It looks as though they have put the minimal amount of solder paste on
the surface and it has not reached the correct fusing temperature for the
required time during reflow.

Can this be anything other than a reflow/paste problem?  This customer
will not accept our conclusion and is looking for compensation.

Rgds

Richard Wood-Roe
Artetch Circuits Ltd
Littlehampton, UK
[log in to unmask]