solder joint problem?

Kim Ji Tae: On your suggested possible causes ...

1) Reflow profile => Could be. What are your profiles? 2) Au finish fail => Could be, but that the connections can be made with hand soldering makes this a low probability cause, assuming the activity of the flux used in hand soldering is not significantly different from that in the paste. 3) IMC => IMC ( intermetalic compound), is unlikely cause, because: � Hand soldering corrects the problem. � Yet, each reflow and hand soldering cycle can only increase intermetalics in SMD connections, which unlike wave soldered connections are not washed away by solder flow.

Alternately, things to consider are: 1 Why solder this 20 pitch Thin (Tape??) Quad Flat Package during the first reflow cycle, rather than the second cycle? 2 Assuming you�ve thought through the reflow sequence, let me pose other possible causes: � TQFP component weight. There�s ample information in the archives on second side component weights. One way to check this would be to apply adhesive on each corner of the TQFP and cure it after the first reflow. � Oven fan pressure is high

reply »

Hi ji tae,

we had a similar problem with QFPs 0,5 pitch processed by one subcontractor. The symptoms we noticed were: - the solderjointshape and wetting looked normal - the surface had a dull gray appearance - by pushing a lead sideways with a tweezer or pin we could move the lead easily with very little force until it separated from the pad ( with a good solderjoint you would normally separate the pad from the PCB by doing this ) - the malfunction of the boards occured during first test or after a couple of days in the systems. - additionell handsoldering solved the problem.

The problem was an improper reflow profile they used and it was suspected that an reoxidation took place due to too much oxygen blown into the joint by their equipment. They never told us what they did actually and I suppose they didn�t know what they where doing wrong. In addition the leads had a gold finish with an undetermined thickness which didn�t make any problems with correct reflow profiles but might have added additional problems with their profiles.

I would agree with Dave according the measures to be taken, check the profile (consider the time the flux stays active), try to do the finepitch in second reflow.

Good luck Wolfgang

reply »

It has been our experience that the problem you have discribed was not solder process related. We had two such instances in the past and both were component related. The first case involved a PLCC28 that failed test. Reflowing the solder seemed to correct the problem when in fact it was just the heat applied that made the circuit function. The problem was component related. The second instance involved a memory device, same symptoms. Turned out the vendor changed the die and the device was more susceptible to noise.

reply »